28 | 
29 |
30 |
31 | #### Step 2: 解压zip包，双击`RapidVideOCR.exe`（以Windows平台为例）
32 |
33 | #### Step 3: 界面如下图所示
34 |
35 |
29 |
36 | 
37 |
38 |
39 | #### Step 4: 界面各个部分介绍
40 |
41 | RapidVideOCR:
42 |
43 | - **图像目录**：指的是 **VideoSubFinder** 软件生成的 **RGBImages** 或者 **TXTImages** 目录，必须是这两个目录之一
44 | - **保存路径**：识别转换后的结果，包括srt文件、ass文件、和txt文件
45 | - **叠图识别**：勾选后，识别速度会变快。准确率可能差些
46 |
47 | VideoSubFinder + RapidVideOCR:
48 |
49 | - **VSF exe路径** ：本地安装的VideoSubFinder全路径，例如:`G:\ProgramFiles\VideoSubFinder_6.10_x64\Release_x64\VideoSubFinderWXW.exe`
50 | - **视频目录** ：想要提取硬字幕的视频所在目录
51 | - **保存路径** ：保存提取结果的目录
52 |
53 | #### Step 5：上述都填好之后，点击OK按钮，即可开始识别
54 |
--------------------------------------------------------------------------------
/docs/blog/posts/rec_other_lan.md:
--------------------------------------------------------------------------------
1 | ---
2 | title: 如何识别其他语言字幕？
3 | date: 2025-09-10
4 | authors: [SWHL]
5 | slug: how-to-rec-other-lan
6 | categories:
7 | - General
8 | comments: true
9 | ---
10 |
11 |
12 | ### 引言
13 |
14 | - 当前，RapidVideOCR是直接使用的`rapidocr_onnxruntime`的默认配置，因此仅能识别中英文的字幕文字。
15 | - 由于`rapidocr_onnxruntime`具备传入其他多语言识别模型的接口，因此RapidVieOCR具备了可扩展性，本篇文章特此来说明如何操作使用。
16 | - 本篇文章以[discussions #40](https://github.com/SWHL/RapidVideOCR/discussions/40)中提出的识别法语字幕为例说明，其他语种同理可得。
17 |
18 | ### 1. 正确安装使用RapidVideOCR
19 |
20 | 请参考教程[link](https://swhl.github.io/RapidVideOCR/docs/tutorial/senior/)
21 |
22 | ### 2. 借助PaddleOCRConvert工具来转换法语识别模型为ONNX
23 |
24 | 请参考教程[link](https://github.com/RapidAI/PaddleOCRModelConvert)
25 |
26 | 其中，
27 |
28 | 模型路径：`https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/french_mobile_v2.0_rec_infer.tar`
29 |
30 | 字典路径：`https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/dygraph/ppocr/utils/dict/french_dict.txt`
31 |
32 | 其他语言的模型下载地址参见：[paddleocr whl](https://files.pythonhosted.org/packages/8f/d0/1a2f9430f61781beb16556182baa938e8f93c8b46c27ad5865a5655fae05/paddleocr-2.7.0.3-py3-none-any.whl)源码中`paddleocr.py`文件中
33 |
34 | 字典链接参见：[link](https://github.com/PaddlePaddle/PaddleOCR/tree/799c144ab3b0b5d19a37c7e85c47e88ff27c643d/ppocr/utils/dict)
35 |
36 | 最终可以得到一个法语识别模型：`french_mobile_v2.0_rec_infer.onnx`
37 |
38 | ### 3. 识别法语字幕
39 |
40 | !!! note
41 |
42 | `rapid_videocr>=v3.0.0`
43 |
44 | ```python linenums="1"
45 | from rapid_videocr import RapidVideOCR, RapidVideOCRInput
46 |
47 | input_args = RapidVideOCRInput(
48 | is_batch_rec=False,
49 | ocr_params={"Rec.model_path": "french_mobile_v2.0_rec_infer.onnx"},
50 | )
51 | extractor = RapidVideOCR(input_args)
52 |
53 | rgb_dir = "test_files/RGBImagesTiny"
54 | save_dir = "outputs"
55 | save_name = "a"
56 |
57 | # outputs/a.srt outputs/a.ass outputs/a.txt
58 | extractor(rgb_dir, save_dir, save_name=save_name)
59 | ```
60 |
--------------------------------------------------------------------------------
/overrides/partials/comments.html:
--------------------------------------------------------------------------------
1 | {% if page.meta.comments %}
2 |
37 | {{ lang.t("meta.comments") }}
3 | 4 | 5 | 10 | 11 | 12 | 13 | 47 | {% endif %} -------------------------------------------------------------------------------- /docs/index.md: -------------------------------------------------------------------------------- 1 | --- 2 | comments: true 3 | hide: 4 | - navigation 5 | - toc 6 | --- 7 | 8 |
9 | 
10 |
11 |
12 | 
13 | 
14 | 
15 | 
16 | 
17 | 
18 | 
19 | 
20 |
21 |
22 |
23 | ### 简介
24 |
25 | `rapid_videocr`是一个自动视频硬字幕提取，生成对应`srt | ass | txt`文件的工具。
26 |
27 | 支持字幕语言：[支持语种列表](https://rapidai.github.io/RapidOCRDocs/main/model_list/#_4)，因为该工具依赖`rapidocr`库，因此`rapidocr`支持识别的语言，`rapid_videocr`均是支持的。
28 |
29 | 优势如下：
30 |
31 | - **提取更快**：与[VideoSubFinder](https://sourceforge.net/projects/videosubfinder/)软件结合使用，提取关键字幕帧更快。
32 | - **识别更准**：采用[RapidOCR](https://github.com/RapidAI/RapidOCR)作为识别库。
33 | - **使用更方便**：pip直接安装即可使用。
34 |
35 | 如果有帮助到您的话，请给个小星星⭐。
36 |
37 | ### 整体框架
38 |
39 | ```mermaid
40 | flowchart LR
41 | A[/Video/] --Extract subtitle key frame--> B(VideoSubFinder) --OCR-->C(RapidVideOCR)
42 | C --Convert--> D[/"SRT | ASS | TXT"/]
43 | ```
44 |
45 | ### [在线Demo](https://huggingface.co/spaces/SWHL/RapidVideOCR)
46 |
47 |
10 |
13 |
14 |
15 |
16 |
17 |
18 |
19 |
20 |
21 |
48 | 
49 |
50 |
--------------------------------------------------------------------------------
/assets/colab-badge.svg:
--------------------------------------------------------------------------------
1 |
2 |
--------------------------------------------------------------------------------
/docs/tutorial/beginner.en.md:
--------------------------------------------------------------------------------
1 | ---
2 | comments: true
3 | hide:
4 | - toc
5 | ---
6 |
7 | ### Introduction
8 |
9 | Considering that most people who extract video subtitles are not programmers, in order to lower the threshold of use, we hereby launch the GUI version of RapidVideOCR Desktop.
10 |
11 | RapidVideOCR Desktop needs to be used with VideoSubFinder. The relationship between the two is shown in the figure below:
12 |
13 | ```mermaid
14 | flowchart LR
15 | A[VideoSubFinder] -- Extract subtitle keyframes --> B[RapidVideOCR] -- OCR --> C(SRT)
16 | ```
17 |
18 | ### [VideoSubFinder Tutorial (in Chinese)](https://blog.csdn.net/shiwanghualuo/article/details/129174857)
19 |
20 | ### RapidVideOCR Desktop Tutorial
21 |
22 | #### Step 1: Download the **RapidVideOCR Desktop** release archive for your corresponding platform
23 |
24 | - [Download the latest version from Github](https://github.com/SWHL/RapidVideOCRDesktop/releases)
25 | - Download from QQ Group shared file, QQ Group number: 706807542, or scan the QR code to join：
26 |
27 |
49 |
28 |
29 |
30 |
31 | #### Step 2: Uncompress the zip archive, and double-click `RapidVideOCR.exe` (using Windows for example)
32 |
33 | #### Step 3: The GUI is shown below
34 |
35 |
29 |
36 |
37 |
38 |
39 | #### Step 4: Introduction to each part of the GUI
40 |
41 | RapidVideOCR:
42 |
43 | - **Image Directory (图像目录)**: Points to the **RGBImages** or **TXTImages** directory generated by **VideoSubFinder**，must be one of the two directories
44 | - **Output Directory (保存路径)**: Where to save the converted results，including srt files, ass files, and txt files
45 | - **Overlay Recognition (叠图识别)**: After checking, the recognition speed will be
46 | faster. The accuracy may be lower
47 |
48 | VideoSubFinder + RapidVideOCR:
49 |
50 | - **VSF exe path**: The full path to the locally installed VideoSubFinder executable, for example: `G:\ProgramFiles\VideoSubFinder_6.10_x64\Release_x64\VideoSubFinderWXW.exe`
51 | - **Video path**: The path to where the video to extract hard subtitles from is located
52 | - **Output directory**：The directory where the extraction results are saved
53 |
54 | #### Step 5: After filling in the above, click the OK button to start recognition
55 |
--------------------------------------------------------------------------------
/setup.py:
--------------------------------------------------------------------------------
1 | # -*- encoding: utf-8 -*-
2 | # @Author: SWHL
3 | # @Contact: liekkaskono@163.com
4 | import sys
5 | from pathlib import Path
6 | from typing import List
7 |
8 | import setuptools
9 | from get_pypi_latest_version import GetPyPiLatestVersion
10 |
11 |
12 | def read_txt(txt_path: str) -> List:
13 | if not isinstance(txt_path, str):
14 | txt_path = str(txt_path)
15 |
16 | with open(txt_path, "r", encoding="utf-8") as f:
17 | data = list(map(lambda x: x.rstrip("\n"), f))
18 | return data
19 |
20 |
21 | def get_readme() -> str:
22 | root_dir = Path(__file__).resolve().parent
23 | readme_path = str(root_dir / "docs" / "doc_whl.md")
24 | with open(readme_path, "r", encoding="utf-8") as f:
25 | readme = f.read()
26 | return readme
27 |
28 |
29 | MODULE_NAME = "rapid_videocr"
30 |
31 | obtainer = GetPyPiLatestVersion()
32 | latest_version = obtainer(MODULE_NAME)
33 | VERSION_NUM = obtainer.version_add_one(latest_version)
34 |
35 | # 优先提取commit message中的语义化版本号，如无，则自动加1
36 | if len(sys.argv) > 2:
37 | match_str = " ".join(sys.argv[2:])
38 | matched_versions = obtainer.extract_version(match_str)
39 | if matched_versions:
40 | VERSION_NUM = matched_versions
41 | sys.argv = sys.argv[:2]
42 |
43 | setuptools.setup(
44 | name=MODULE_NAME,
45 | version=VERSION_NUM,
46 | platforms="Any",
47 | description="Tool for extracting hard subtitles from videos.",
48 | long_description=get_readme(),
49 | long_description_content_type="text/markdown",
50 | author="SWHL",
51 | author_email="liekkaskono@163.com",
52 | url="https://github.com/SWHL/RapidVideOCR.git",
53 | license="Apache-2.0",
54 | include_package_data=True,
55 | install_requires=read_txt("requirements.txt"),
56 | packages=setuptools.find_packages(),
57 | keywords=["rapidocr,videocr,subtitle"],
58 | classifiers=[
59 | "Programming Language :: Python :: 3.6",
60 | "Programming Language :: Python :: 3.7",
61 | "Programming Language :: Python :: 3.8",
62 | "Programming Language :: Python :: 3.9",
63 | "Programming Language :: Python :: 3.10",
64 | "Programming Language :: Python :: 3.11",
65 | "Programming Language :: Python :: 3.12",
66 | "Programming Language :: Python :: 3.13",
67 | ],
68 | python_requires=">=3.6",
69 | entry_points={
70 | "console_scripts": [f"{MODULE_NAME}={MODULE_NAME}.main:main"],
71 | },
72 | )
73 |
--------------------------------------------------------------------------------
/docs/blog/posts/rec_other_lan.en.md:
--------------------------------------------------------------------------------
1 | ---
2 | title: How to recognize other language subtitle?
3 | date: 2025-09-10
4 | authors: [SWHL]
5 | slug: how-to-rec-other-lan
6 | categories:
7 | - General
8 | comments: true
9 |
10 | ---
11 | ### Introduction
12 |
13 | Currently, RapidVideOCR directly uses the default configuration of `rapidocr_onnxruntime`, so it can only do OCR for subtitles in Chinese and English.
14 |
15 | Since `rapidocr_onnxruntime` has an interface for passing in other multilingual recognition models, RapidVieOCR has scalability. This article is here to explain how to use it.
16 |
17 | This article takes the French OCR solution proposed in [discussions #40](https://github.com/SWHL/RapidVideOCR/discussions/40) as an example, and other languages can be done in the same way.
18 |
19 | ### 1. Correctly install and use RapidVideOCR
20 |
21 | Please refer to this [link](https://swhl.github.io/RapidVideOCR/en/docs/tutorial/senior/)
22 |
23 | ### 2. Use PaddleOCR Convert tool to convert French recognition model to ONNX
24 |
25 | Please refer to the tutorial [link](https://github.com/RapidAI/PaddleOCRModelConvert)
26 |
27 | Using:
28 |
29 | Model path: `https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/french_mobile_v2.0_rec_infer.tar`,
30 |
31 | Dictionary path: `https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/dygraph/ppocr/utils/dict/french_dict.txt`
32 |
33 | For model download links for other languages, please refer to: [paddleocr whl](https://files.pythonhosted.org/packages/8f/d0/1a2f9430f61781beb16556182baa938e8f93c8b46c27ad5865a5655fae05/paddleocr-2.7.0.3-py3-none-any.whl) in the source `paddleocr.py` file
34 |
35 | For dictionary models, see: [link](https://github.com/PaddlePaddle/PaddleOCR/tree/799c144ab3b0b5d19a37c7e85c47e88ff27c643d/ppocr/utils/dict)
36 |
37 | Finally, a French recognition model can be obtained: `french_mobile_v2.0_rec_infer.onnx`
38 |
39 | ### 3. OCR French subtitles
40 |
41 | !!! note
42 |
43 | `rapid_videocr>=v3.0.0`
44 |
45 | ```python linenums="1"
46 | from rapid_videocr import RapidVideOCR, RapidVideOCRInput
47 |
48 | input_args = RapidVideOCRInput(
49 | is_batch_rec=False,
50 | ocr_params={"Rec.model_path": "french_mobile_v2.0_rec_infer.onnx"},
51 | )
52 | extractor = RapidVideOCR(input_args)
53 |
54 | rgb_dir = "test_files/RGBImagesTiny"
55 | save_dir = "outputs"
56 | save_name = "a"
57 |
58 | # outputs/a.srt outputs/a.ass outputs/a.txt
59 | extractor(rgb_dir, save_dir, save_name=save_name)
60 | ```
61 |
--------------------------------------------------------------------------------
/rapid_videocr/vsf_cli.py:
--------------------------------------------------------------------------------
1 | # -*- encoding: utf-8 -*-
2 | # @Author: SWHL
3 | # @Contact: liekkaskono@163.com
4 | import subprocess
5 | from dataclasses import asdict, dataclass
6 | from typing import Optional
7 |
8 |
9 | @dataclass
10 | class VideoSubFinderInput:
11 | vsf_exe_path: str
12 | clear_dirs: bool = True
13 | run_search: bool = True
14 | create_cleared_text_images: bool = True
15 | create_empty_sub: Optional[str] = None
16 | create_sub_from_cleared_txt_images: Optional[str] = None
17 | create_sub_from_txt_results: Optional[str] = None
18 | open_video_opencv: bool = True
19 | open_video_ffmpeg: bool = False
20 | use_cuda: bool = False
21 | start_time: Optional[str] = None
22 | end_time: Optional[str] = None
23 | top_video_image_percent_end: float = 0.2
24 | bottom_video_image_percent_end: float = 0.0
25 | left_video_image_percent_end: float = 0.0
26 | right_video_image_percent_end: float = 1.0
27 | general_settings: Optional[str] = None
28 | num_threads: int = 2
29 | num_ocr_threads: int = 1
30 |
31 |
32 | class VideoSubFinder:
33 | SHORT_FLAG_MAP = {
34 | "clear_dirs": "-c",
35 | "run_search": "-r",
36 | "create_cleared_text_images": "-ccti",
37 | "create_empty_sub": "-ces",
38 | "create_sub_from_cleared_txt_images": "-cscti",
39 | "create_sub_from_txt_results": "-cstxt",
40 | "open_video_opencv": "-ovocv",
41 | "open_video_ffmpeg": "-ovffmpeg",
42 | "use_cuda": "-uc",
43 | "start_time": "-s",
44 | "end_time": "-e",
45 | "top_video_image_percent_end": "-te",
46 | "bottom_video_image_percent_end": "-be",
47 | "left_video_image_percent_end": "-le",
48 | "right_video_image_percent_end": "-re",
49 | "general_settings": "-gs",
50 | "num_threads": "-nthr",
51 | "num_ocr_threads": "-nocrthr",
52 | }
53 |
54 | def __init__(self, input_params: VideoSubFinderInput):
55 | param_dict = asdict(input_params)
56 | run_list = [input_params.vsf_exe_path]
57 | for k, v in param_dict.items():
58 | if k == "vsf_exe_path":
59 | continue
60 |
61 | if v is None or str(v) == "False":
62 | continue
63 |
64 | flag = self.SHORT_FLAG_MAP[k]
65 | run_list.append(f"{flag}" if str(v) == "True" else f"{flag} {v}")
66 | self.run_list = run_list
67 |
68 | def __call__(self, video_path: str, output_dir: str) -> str:
69 | self.run_list.extend(["--input_video", video_path, "--output_dir", output_dir])
70 | try:
71 | subprocess.run(self.run_list, check=False)
72 | return output_dir
73 | except Exception as e:
74 | raise e
75 |
--------------------------------------------------------------------------------
/docs/index.en.md:
--------------------------------------------------------------------------------
1 | ---
2 | comments: true
3 | hide:
4 | - navigation
5 | - toc
6 | ---
7 |
8 |
37 |
9 |
10 |
11 |
12 |
13 | 
14 |
15 | 
16 |
17 |
18 |
19 |
20 |
21 |
22 |
23 |
24 |
25 | ### Introduction
26 |
27 | - Video hard subtitle extraction, automatically generate the corresponding `srt | ass | txt` file.
28 | - Supported subtitle languages: Chinese | English (For other supported languages, see: [List of supported languages](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_ch/multi_languages.md#%E8%AF%AD%E7%A7%8D%E7%BC%A9%E5%86%99))
29 | - The advantages are as follows:
30 | - **Faster extraction**: Used in conjunction with [VideoSubFinder](https://sourceforge.net/projects/videosubfinder/) software to extract key subtitle frames faster.
31 | - **More accurate recognition**: Use [RapidOCR](https://github.com/RapidAI/RapidOCR) as the recognition library.
32 | - **More convenient to use**: pip can be installed directly and used.
33 |
34 | - For desktop EXE version, please go to [RapidVideOCRDesktop](https://github.com/SWHL/RapidVideOCRDesktop).
35 | - If it helps you, please give a star ⭐.
36 |
37 | ### Overall framework
38 |
39 | ```mermaid
40 | flowchart LR
41 | A[/Video/] --Extract subtitle key frame--> B(VideoSubFinder) --OCR-->C(RapidVideOCR)
42 | C --Convert--> D[/"SRT | ASS | TXT"/]
43 | ```
44 |
45 | ### [Online Demo](https://huggingface.co/spaces/SWHL/RapidVideOCR)
46 |
47 |
10 |
11 |
14 |
15 |
16 |
17 |
18 |
19 |
20 |
21 |
22 |
23 |
48 |
49 |
50 |
--------------------------------------------------------------------------------
/.gitignore:
--------------------------------------------------------------------------------
1 | rapid_videocr/video_sub_finder/libs
2 | *.vscode
3 | outputs
4 |
5 | *.pyc
6 |
7 | *.onnx
8 |
9 | temp/
10 | test_files/
11 |
12 | .DS_Store
13 |
14 | *.bin
15 |
16 | .mypy_cache
17 |
18 | # Created by .ignore support plugin (hsz.mobi)
19 | ### Python template
20 | # Byte-compiled / optimized / DLL files
21 | __pycache__/
22 | *.py[cod]
23 | *$py.class
24 | .pytest_cache
25 |
26 | # C extensions
27 | *.so
28 |
29 | # Distribution / packaging
30 | .Python
31 | build/
32 | develop-eggs/
33 | dist/
34 | downloads/
35 | eggs/
36 | .eggs/
37 | lib/
38 | lib64/
39 | parts/
40 | sdist/
41 | var/
42 | wheels/
43 | pip-wheel-metadata/
44 | share/python-wheels/
45 | *.egg-info/
46 | .installed.cfg
47 | *.egg
48 | MANIFEST
49 |
50 | # PyInstaller
51 | # Usually these files are written by a python script from a template
52 | # before PyInstaller builds the exe, so as to inject date/other infos into it.
53 | # *.manifest
54 | # *.spec
55 | *.res
56 |
57 | # Installer logs
58 | pip-log.txt
59 | pip-delete-this-directory.txt
60 |
61 | # Unit test / coverage reports
62 | htmlcov/
63 | .tox/
64 | .nox/
65 | .coverage
66 | .coverage.*
67 | .cache
68 | nosetests.xml
69 | coverage.xml
70 | *.cover
71 | *.py,cover
72 | .hypothesis/
73 | .pytest_cache/
74 |
75 | # Translations
76 | *.mo
77 | *.pot
78 |
79 | # Django stuff:
80 | *.log
81 | local_settings.py
82 | db.sqlite3
83 | db.sqlite3-journal
84 |
85 | # Flask stuff:
86 | instance/
87 | .webassets-cache
88 |
89 | # Scrapy stuff:
90 | .scrapy
91 |
92 | # Sphinx documentation
93 | docs/_build/
94 |
95 | # PyBuilder
96 | target/
97 |
98 | # Jupyter Notebook
99 | .ipynb_checkpoints
100 |
101 | # IPython
102 | profile_default/
103 | ipython_config.py
104 |
105 | # pyenv
106 | .python-version
107 |
108 | # pipenv
109 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
110 | # However, in case of collaboration, if having platform-specific dependencies or dependencies
111 | # having no cross-platform support, pipenv may install dependencies that don't work, or not
112 | # install all needed dependencies.
113 | #Pipfile.lock
114 |
115 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow
116 | __pypackages__/
117 |
118 | # Celery stuff
119 | celerybeat-schedule
120 | celerybeat.pid
121 |
122 | # SageMath parsed files
123 | *.sage.py
124 |
125 | # Environments
126 | .env
127 | .venv
128 | env/
129 | venv/
130 | ENV/
131 | env.bak/
132 | venv.bak/
133 |
134 | # Spyder project settings
135 | .spyderproject
136 | .spyproject
137 |
138 | # Rope project settings
139 | .ropeproject
140 |
141 | # mkdocs documentation
142 | /site
143 |
144 | # mypy
145 | .mypy_cache/
146 | .dmypy.json
147 | dmypy.json
148 |
149 | # Pyre type checker
150 | .pyre/
151 |
152 | #idea
153 | .vs
154 | .vscode
155 | .idea
156 | /images
157 | /models
158 |
159 | #models
160 | *.onnx
161 |
162 | *.ttf
163 | *.ttc
164 |
165 | long1.jpg
166 |
167 | *.bin
168 | *.mapping
169 | *.xml
170 |
171 | *.pdiparams
172 | *.pdiparams.info
173 | *.pdmodel
174 |
175 | .DS_Store
--------------------------------------------------------------------------------
/rapid_videocr/export.py:
--------------------------------------------------------------------------------
1 | # -*- encoding: utf-8 -*-
2 | # @Author: SWHL
3 | # @Contact: liekkaskono@163.com
4 | from abc import ABC, abstractmethod
5 | from pathlib import Path
6 | from typing import List
7 |
8 | from .utils.typings import OutputFormat
9 | from .utils.utils import write_txt
10 |
11 |
12 | class ExportStrategy(ABC):
13 | @abstractmethod
14 | def export(
15 | self,
16 | save_dir: Path,
17 | save_name: str,
18 | srt_result: List[str],
19 | ass_result: List[str],
20 | txt_result: List[str],
21 | ):
22 | pass
23 |
24 |
25 | class TxtExportStrategy(ExportStrategy):
26 | def export(
27 | self,
28 | save_dir: Path,
29 | save_name: str,
30 | srt_result: List[str],
31 | ass_result: List[str],
32 | txt_result: List[str],
33 | ):
34 | file_path = save_dir / f"{save_name}.txt"
35 | write_txt(file_path, txt_result)
36 |
37 |
38 | class SrtExportStrategy(ExportStrategy):
39 | def export(
40 | self,
41 | save_dir: Path,
42 | save_name: str,
43 | srt_result: List[str],
44 | ass_result: List[str],
45 | txt_result: List[str],
46 | ):
47 | file_path = save_dir / f"{save_name}.srt"
48 | write_txt(file_path, srt_result)
49 |
50 |
51 | class AssExportStrategy(ExportStrategy):
52 | def export(
53 | self,
54 | save_dir: Path,
55 | save_name: str,
56 | srt_result: List[str], # unused here but kept for signature
57 | ass_result: List[str],
58 | txt_result: List[str],
59 | ):
60 | header = [
61 | "[Script Info]",
62 | "; Script generated by RapidVideOCR",
63 | "ScriptType: v4.00+",
64 | "PlayResX: 1920",
65 | "PlayResY: 1080",
66 | "",
67 | "[V4+ Styles]",
68 | "Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, "
69 | "Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, "
70 | "Alignment, MarginL, MarginR, MarginV, Encoding",
71 | "Style: Default,Arial,54,&H00FFFFFF,&H0000FFFF,&H00000000,&H64000000,0,0,0,0,100,100,0,0,1,2,2,2,10,10,10,1",
72 | "",
73 | "[Events]",
74 | "Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text",
75 | ]
76 |
77 | file_path = save_dir / f"{save_name}.ass"
78 | write_txt(file_path, header + [""] + ass_result)
79 |
80 |
81 | class AllExportStrategy(ExportStrategy):
82 | def export(
83 | self,
84 | save_dir: Path,
85 | save_name: str,
86 | srt_result: List[str],
87 | ass_result: List[str],
88 | txt_result: List[str],
89 | ):
90 | txt_export = TxtExportStrategy()
91 | srt_export = SrtExportStrategy()
92 | ass_export = AssExportStrategy()
93 |
94 | txt_export.export(save_dir, save_name, srt_result, ass_result, txt_result)
95 | srt_export.export(save_dir, save_name, srt_result, ass_result, txt_result)
96 | ass_export.export(save_dir, save_name, srt_result, ass_result, txt_result)
97 |
98 |
99 | class ExportStrategyFactory:
100 | @staticmethod
101 | def create_strategy(out_format: str = OutputFormat.ALL.value) -> ExportStrategy:
102 | strategies = {
103 | OutputFormat.TXT.value: TxtExportStrategy(),
104 | OutputFormat.SRT.value: SrtExportStrategy(),
105 | OutputFormat.ASS.value: AssExportStrategy(),
106 | OutputFormat.ALL.value: AllExportStrategy(),
107 | }
108 |
109 | if strategy := strategies.get(out_format):
110 | return strategy
111 | raise ValueError(f"Unsupported output format: {out_format}")
112 |
--------------------------------------------------------------------------------
/docs/changelog.md:
--------------------------------------------------------------------------------
1 | ---
2 | comments: true
3 | hide:
4 | - toc
5 | ---
6 |
7 |
8 | ### 📣 后续更新日志将移步到[Release](https://github.com/SWHL/RapidVideOCR/releases)界面，这里不再更新
9 |
10 | #### 🚩2023-10-08 v2.2.8 update
11 |
12 | - 适配`rapidocr_onnxruntime`的相关参数，可以通过RapidVideOCR类传入，从而更加灵活指定不同语言的模型。
13 |
14 | #### ♦ 2023-08-05 v2.2.4 update
15 |
16 | - 修复批量识别模式下，索引错误。
17 | - 添加日志记录模块，便于使用桌面版，快速记录问题，便于反馈。
18 |
19 | #### 🛶2023-07-19 v2.2.3 update
20 |
21 | - 增加对VSF的参数的适配，命令行模式和类初始化时，可以指定VSF命令的同名参数。详细使用参见[link](https://github.com/SWHL/RapidVideOCR/wiki/RapidVideOCR%E9%AB%98%E7%BA%A7%E6%95%99%E7%A8%8B%EF%BC%88%E6%9C%89python%E5%9F%BA%E7%A1%80%E7%9A%84%E5%B0%8F%E4%BC%99%E4%BC%B4%EF%BC%89)
22 |
23 | #### 🤓2023-07-08 v2.2.2 update
24 |
25 | - 修复批量识别时，不能读取中文路径的问题
26 | - 修复漏轴时，SRT中跳过问题。目前当出现某一轴未能识别，则会空出位置，便于校对。
27 | - 保留VSF识别的中间结果
28 |
29 | #### 🐲2023-06-22 v2.2.0 update
30 |
31 | - 该版本是向`v2.1.x`兼容的，也就是之前用法依然可以。
32 | - 将VSF的CLI整合到库中，只需指定`VideoSubFinderWXW.exe`的全路径即可。
33 | - 增加批量识别功能，指定视频目录，即可自动提取目录下所有视频字幕
34 | - 使用示例, 参见：[demo.py](https://github.com/SWHL/RapidVideOCR/blob/main/demo.py)
35 |
36 | #### 😀2023-05-12 v2.1.7 update
37 |
38 | - 优化代码
39 | - 添加`save_name`参数，可以灵活指定保存的`srt | txt`文件名称，默认是`result`
40 |
41 | #### 🐱2023-03-27 v2.1.6 update
42 |
43 | - 修复时间轴对不齐问题，详情参见[issue 23](https://github.com/SWHL/RapidVideOCR/issues/23)
44 |
45 | #### 👽2023-03-23 v2.1.5 update
46 |
47 | - 添加打印到屏幕的控制参数`is_print_console`
48 | - 调整`out_format`参数位置到初始化类时
49 |
50 | #### 😀2023-03-14 v2.1.3 update
51 |
52 | - 修复输入`TXTImages`目录且叠字识别时错误
53 |
54 | #### 😜2023-03-12 v2.1.2 update
55 |
56 | - 修复索引错误，[#22](https://github.com/SWHL/RapidVideOCR/issues/22)
57 |
58 | #### 🎢2023-03-11 v2.1.1 update
59 |
60 | - 修复单图识别与之前版本差异问题
61 | - 默认识别模式更改为单图识别，是否使用叠图识别，请自行决定
62 |
63 | #### 🥇2023-03-10 v2.1.0 update
64 |
65 | - 添加叠字识别功能，速度更快，默认是叠字识别功能
66 |
67 | #### 🎈2023-03-02 v2.0.5~7 update
68 |
69 | - 修复生成的srt文件中的格式错误， [#19](https://github.com/SWHL/RapidVideOCR/issues/19)
70 |
71 | #### 🎫2023-02-17 v2.0.4 update
72 |
73 | - 针对传入的`TXTImages`目录，作了优化处理。相比于传入`RGBImages`，会更快和更准。推荐传入`TXTImages`目录
74 |
75 | #### 💎2023-02-17 v2.0.2 update
76 |
77 | - 修复同行字幕识别丢失空格问题
78 |
79 | #### 🎈2023-01-29 v1.1.10 update
80 |
81 | - 修复帧索引转时间戳时，索引为空错误
82 |
83 | #### 🧨2023-01-28 v1.1.9 update
84 |
85 | - 修复时间轴对不齐问题
86 |
87 | #### 👊 2023-01-15 v1.1.4 update
88 |
89 | - 添加输出txt格式的选项，目前v1.1.4版本默认输出srt和txt两种格式
90 | - 添加根据运行程序屏幕大小，调节选择字幕的框大小
91 |
92 | #### 🌈2023-01-10 v1.0.3 update
93 |
94 | - 将decord替换为OpenCV，因为decord处理MP4时，存在内存泄漏问题。详情参见：[#208](https://github.com/dmlc/decord/issues/208)
95 |
96 | #### 🎄2022-12-04 update
97 |
98 | - 添加交互式框定字幕位置功能，默认开启，更加好用，详情可参考下面的GIF图。感谢@[Johndirr](https://github.com/Johndirr)的建议。
99 | - 优化代码结构，将RapidOCR相关模型和配置文件放到`rapidocr`目录下
100 | - `rapidvideocr`的配置文件也放到对应目录下
101 |
102 | #### 🌼2022-05-08 update
103 |
104 | - 添加交互式确定二值化字幕图像阈值操作，仅仅支持Windows系统，可以通过`is_select_threshold = True`来使用
105 | - 优化代码
106 |
107 | #### 🎉2022-05-03 update
108 |
109 | - 添加GPU支持，具体配置教程参见：[onnxruntime-gpu版推理配置](https://github.com/RapidAI/RapidOCR/blob/main/python/onnxruntime_infer/README.md#onnxruntime-gpu%E7%89%88%E6%8E%A8%E7%90%86%E9%85%8D%E7%BD%AE)
110 | - 添加日文的支持，可以支持更多语种，具体参见：[支持语种列表](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_ch/multi_languages.md#%E8%AF%AD%E7%A7%8D%E7%BC%A9%E5%86%99)
111 |
112 | #### 💡2022-05-01 update
113 |
114 | - 添加语音模块部分位于分支`asr_module`
115 | - 添加语音识别模块，由于该模块中解码部分只能在Linux和Mac上运行，因此如果想要使用该模块，请在Linux和Mac上。
116 | - 目前语音识别代码来自[RapidASR/python](https://github.com/RapidAI/RapidASR/tree/main/python/base_paddlespeech)部分。模型来自[PaddleSpeech](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/aishell/asr0)
117 | - 经过简单测试，语音识别模块不是太准。-_-!
118 |
119 | #### 2022-03-09 update
120 |
121 | - 添加[常见问题模块](./faq.md)，可以帮助大家跳过常见的小问题
122 |
123 | #### 2021-12-14 update
124 |
125 | - [x] 背景去除效果不好，导致丢失某些帧
126 | - 尝试采用图像分割的方法，经过测试，CPU下推理速度太慢，暂时放弃
127 | - 目前采用的固定的二值化阈值
128 | - [x] (2021-12-14)完善对应的英文文档
129 | - [x] (2021-12-14)添加运行耗时基准
130 | - [x] 添加具体参数说明
131 | - [x] 制作项目Logo
132 | - [ ] 更多的测试
133 |
--------------------------------------------------------------------------------
/docs/README_zh.md:
--------------------------------------------------------------------------------
1 |
49 |
2 |
3 |
4 |
5 |
6 | 
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
16 | 简体中文 | [English](https://github.com/SWHL/RapidVideOCR)
17 |
18 |
19 | ### 简介
20 |
21 | `rapid_videocr`是一个自动视频硬字幕提取，生成对应`srt | ass | txt`文件的工具。
22 |
23 | 支持字幕语言：[支持语种列表](https://rapidai.github.io/RapidOCRDocs/main/model_list/#_4)，因为该工具依赖`rapidocr`库，因此`rapidocr`支持识别的语言，`rapid_videocr`均是支持的。
24 |
25 | 优势如下：
26 |
27 | - **提取更快**：与[VideoSubFinder](https://sourceforge.net/projects/videosubfinder/)软件结合使用，提取关键字幕帧更快。
28 | - **识别更准**：采用[RapidOCR](https://github.com/RapidAI/RapidOCR)作为识别库。
29 | - **使用更方便**：pip直接安装即可使用。
30 |
31 | 桌面EXE版，请移步[RapidVideOCRDesktop](https://github.com/SWHL/RapidVideOCRDesktop)
32 |
33 | 如果有帮助到您的话，请给个小星星⭐。
34 |
35 | ### [在线Demo](https://www.modelscope.cn/studios/liekkas/RapidVideOCR/summary)
36 |
37 |
3 |
4 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
16 | 简体中文 | [English](https://github.com/SWHL/RapidVideOCR)
17 |
38 |
39 |
40 |
41 | ### 整体框架
42 |
43 | ```mermaid
44 | flowchart LR
45 | A[/Video/] --Extract subtitle key frame--> B(VideoSubFinder) --OCR-->C(RapidVideOCR)
46 | C --Convert--> D[/"SRT | ASS | TXT"/]
47 | ```
48 |
49 | ### 安装
50 |
51 | ```bash
52 | pip install rapid_videocr
53 | ```
54 |
55 | ### 使用
56 |
57 | > [!NOTE]
58 | >
59 | > `rapid_videocr`输入图像路径必须是**VideoSubFinder**软件输出的RGBImages或TXTImages的路径。
60 |
61 | ```bash
62 | rapid_videocr -i RGBImages
63 | ```
64 |
65 | 或者python脚本：
66 |
67 | ```python
68 | from rapid_videocr import RapidVideOCR, RapidVideOCRInput
69 |
70 | input_args = RapidVideOCRInput(is_batch_rec=False)
71 | extractor = RapidVideOCR(input_args)
72 |
73 | rgb_dir = "tests/test_files/RGBImages"
74 | save_dir = "outputs"
75 | save_name = "a"
76 |
77 | # outputs/a.srt outputs/a.ass outputs/a.txt
78 | extractor(rgb_dir, save_dir, save_name=save_name)
79 | ```
80 |
81 | ### 文档
82 |
83 | 完整文档请移步：[docs](https://swhl.github.io/RapidVideOCR/)
84 |
85 | ### 贡献者
86 |
87 |
39 |
88 |
89 |
90 |
91 |
105 | 
106 |
107 |
108 | 扫码加入QQ群（706807542）：
109 |
106 |
110 |
111 |
112 |
113 | ### [赞助](https://swhl.github.io/RapidVideOCR/docs/sponsor/)
114 |
115 | 如果您想要赞助该项目，可直接点击当前页最上面的Sponsor按钮，请写好备注( **您的Github账号名称** )，方便添加到赞助列表中。
116 |
117 | ### 开源许可证
118 |
119 | 该项目采用 [Apache 2.0 license](../LICENSE) 开源许可证。
120 |
--------------------------------------------------------------------------------
/mkdocs.yml:
--------------------------------------------------------------------------------
1 | site_name: RapidVideOCR 文档
2 | site_url: https://swhl.github.io/RapidVideOCR/
3 | site_author: SWHL
4 | site_description:
5 | 🎦 Extract video hard subtitles and automatically generate corresponding srt files.
6 |
7 | repo_name: SWHL/RapidVideOCR
8 | repo_url: https://github.com/SWHL/RapidVideOCR
9 |
10 | copyright: Copyright © 2025 Maintained by SWHL.
11 |
12 | theme:
13 | name: material
14 | language: zh
15 | custom_dir: overrides
16 | features:
17 | - announce.dismiss
18 | - content.tooltips
19 | - content.code.copy
20 | - content.tabs.link
21 | - content.footnote.tooltips
22 | - navigation.expand # 默认打开所有的子节
23 | - navigation.tabs # 顶级索引被作为tab
24 | - navigation.tabs.sticky # tab始终可见
25 | - navigation.top # 开启顶部导航栏
26 | - navigation.tracking # 导航栏跟踪
27 | - navigation.footer
28 | - navigation.indexes
29 | - search.highlight # 搜索高亮
30 | - search.share # 搜索分享
31 | - search.suggest # 搜索建议
32 | - toc.follow # 目录跟踪-页面右侧的小目录
33 |
34 | palette:
35 | - media: "(prefers-color-scheme)"
36 | toggle:
37 | icon: material/brightness-auto
38 | name: Switch to light mode
39 |
40 | - media: "(prefers-color-scheme: light)"
41 | scheme: default
42 | toggle:
43 | icon: material/brightness-7
44 | name: Switch to dark mode
45 |
46 | - media: "(prefers-color-scheme: dark)"
47 | scheme: slate
48 | toggle:
49 | icon: material/brightness-4
50 | name: Switch to system preference
51 |
52 | icon:
53 | logo: logo
54 | previous: fontawesome/solid/angle-left
55 | next: fontawesome/solid/angle-right
56 | repo: fontawesome/brands/github
57 | tag:
58 | default-tag: fontawesome/solid/tag
59 | hardware-tag: fontawesome/solid/microchip
60 | software-tag: fontawesome/solid/laptop-code
61 |
62 | plugins:
63 | - tags
64 | - search:
65 | separator: '[\s\u200b\-_,:!=\[\]()"`/]+|\.(?!\d)|&[lg]t;|(?!\b)(?=[A-Z][a-z])'
66 | - i18n:
67 | docs_structure: suffix
68 | fallback_to_default: true
69 | reconfigure_material: true
70 | reconfigure_search: true
71 | languages:
72 | - locale: zh
73 | name: 简体中文
74 | default: true
75 | build: true
76 | - locale: en
77 | name: English
78 | site_name: RapidVideOCR Documentation
79 | link: /en/
80 | nav_translations:
81 | 概览: Overview
82 | 快速开始: Quick Start
83 | 在线demo: Online Demo
84 | 使用教程: Tutorial
85 | 初级教程: Beginner Tutorial
86 | 中级教程: Intermediate Tutorial
87 | 高级教程: Senior Tutorial
88 | 博客: Blog
89 | 如何识别其他语言字幕？: How to recognize other language subtitle?
90 | 常见问题: FAQ
91 | 赞助: Sponsor
92 | 更新日志: Changelog
93 | - git-committers:
94 | repository: SWHL/RapidVideOCR
95 | branch: main
96 | token: !!python/object/apply:os.getenv ["MKDOCS_GIT_COMMITTERS_APIKEY"]
97 | - git-revision-date-localized:
98 | enable_creation_date: true
99 |
100 |
101 | markdown_extensions:
102 | - abbr
103 | - attr_list
104 | - pymdownx.snippets
105 | - pymdownx.critic
106 | - pymdownx.caret
107 | - pymdownx.keys
108 | - pymdownx.mark
109 | - pymdownx.tilde
110 | - footnotes
111 | - def_list
112 | - md_in_html
113 | - tables
114 | - pymdownx.tasklist:
115 | custom_checkbox: true
116 | - toc:
117 | permalink: true
118 | - pymdownx.betterem:
119 | smart_enable: all
120 | - pymdownx.superfences:
121 | custom_fences:
122 | - name: mermaid
123 | class: mermaid
124 | format: !!python/name:pymdownx.superfences.fence_code_format
125 | - pymdownx.emoji:
126 | emoji_index: !!python/name:material.extensions.emoji.twemoji
127 | emoji_generator: !!python/name:material.extensions.emoji.to_svg
128 | - pymdownx.highlight:
129 | anchor_linenums: true
130 | line_spans: __span
131 | pygments_lang_class: true
132 | - pymdownx.inlinehilite
133 | - pymdownx.tabbed:
134 | alternate_style: true
135 | - admonition
136 | - pymdownx.details
137 |
138 | extra:
139 | version:
140 | provider: mike
141 |
142 | nav:
143 | - 概览: index.md
144 | - 快速开始: quickstart.md
145 | - 在线demo: online_demo.md
146 | - 使用教程:
147 | - 初级教程: tutorial/beginner.md
148 | - 中级教程: tutorial/intermediate.md
149 | - 高级教程: tutorial/senior.md
150 | - 常见问题: faq.md
151 | - 博客:
152 | - 如何识别其他语言字幕？: blog/posts/rec_other_lan.md
153 | - 赞助: sponsor.md
154 | - 更新日志: changelog.md
155 |
--------------------------------------------------------------------------------
/cliff.toml:
--------------------------------------------------------------------------------
1 | # git-cliff ~ configuration file
2 | # https://git-cliff.org/docs/configuration
3 |
4 | [changelog]
5 | # A Tera template to be rendered as the changelog's footer.
6 | # See https://keats.github.io/tera/docs/#introduction
7 | # header = """
8 | # # Changelog\n
9 | # All notable changes to this project will be documented in this file. See [conventional commits](https://www.conventionalcommits.org/) for commit guidelines.\n
10 | # """
11 | # A Tera template to be rendered for each release in the changelog.
12 | # See https://keats.github.io/tera/docs/#introduction
13 | body = """
14 | {% for group, commits in commits | group_by(attribute="group") %}
15 | ### {{ group | striptags | trim | upper_first }}
16 | {% for commit in commits
17 | | filter(attribute="scope")
18 | | sort(attribute="scope") %}
19 | - **({{commit.scope}})**{% if commit.breaking %} [**breaking**]{% endif %} \
20 | {{ commit.message }} by [@{{ commit.author.name }}](https://github.com/{{ commit.author.name }}) in [{{ commit.id | truncate(length=7, end="") }}]($REPO/commit/{{ commit.id }})
21 | {%- endfor -%}
22 | {% raw %}\n{% endraw %}\
23 | {%- for commit in commits %}
24 | {%- if commit.scope -%}
25 | {% else -%}
26 | - {% if commit.breaking %} [**breaking**]{% endif %}\
27 | {{ commit.message }} by [@{{ commit.author.name }}](https://github.com/{{ commit.author.name }}) in [{{ commit.id | truncate(length=7, end="") }}]($REPO/commit/{{ commit.id }})
28 | {% endif -%}
29 | {% endfor -%}
30 | {% endfor %}
31 |
32 |
33 | {% if github.contributors | length > 0 %}
34 | ### 🎉 Contributors
35 |
36 | {% for contributor in github.contributors %}
37 | - [@{{ contributor.username }}](https://github.com/{{ contributor.username }})
38 | {%- endfor -%}
39 | {% endif %}
40 |
41 |
42 | {% if version %}
43 | {% if previous.version %}\
44 | **Full Changelog**: [{{ version | trim_start_matches(pat="v") }}]($REPO/compare/{{ previous.version }}..{{ version }})
45 | {% else %}\
46 | **Full Changelog**: [{{ version | trim_start_matches(pat="v") }}]
47 | {% endif %}\
48 | {% else %}\
49 | ## [unreleased]
50 | {% endif %}
51 | """
52 | # A Tera template to be rendered as the changelog's footer.
53 | # See https://keats.github.io/tera/docs/#introduction
54 |
55 | footer = """
56 |
57 | """
58 |
59 | # Remove leading and trailing whitespaces from the changelog's body.
60 | trim = true
61 | # postprocessors
62 | postprocessors = [
63 | # Replace the placeholder `
111 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
16 | [简体中文](https://github.com/SWHL/RapidVideOCR/blob/main/docs/README_zh.md) | English
17 |
18 |
19 | ### Introduction
20 |
21 | `rapid_videocr` is a tool designed for the automatic extraction of hard subtitles from videos and generates corresponding `srt | ass | txt` files.
22 |
23 | Supported subtitle languages: [List of Supported Languages](https://rapidai.github.io/RapidOCRDocs/main/model_list/#_4). Since this tool relies on the `rapidocr` library, any language supported by `rapidocr` for recognition is also supported by `rapid_videocr`.
24 |
25 | Its advantages include:
26 |
27 | - **Faster Extraction**: When used in conjunction with the [VideoSubFinder](https://sourceforge.net/projects/videosubfinder/) software, it can extract key subtitle frames more quickly.
28 | - **More Accurate Recognition**: It utilizes [RapidOCR](https://github.com/RapidAI/RapidOCR) as its recognition library.
29 | - **Easier to Use**: It can be installed directly via pip and is ready to use.
30 |
31 | For the desktop EXE version, please visit [RapidVideOCRDesktop](https://github.com/SWHL/RapidVideOCRDesktop).
32 |
33 | If this tool has been helpful to you, please give it a star ⭐.
34 |
35 | ### [Online Demo](https://huggingface.co/spaces/SWHL/RapidVideOCR)
36 |
37 |
3 |
4 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
16 | [简体中文](https://github.com/SWHL/RapidVideOCR/blob/main/docs/README_zh.md) | English
17 |
38 |
39 |
40 |
41 | ### Overall framework
42 |
43 | ```mermaid
44 | flowchart LR
45 | A[/Video/] --Extract subtitle key frame--> B(VideoSubFinder) --OCR-->C(RapidVideOCR)
46 | C --Convert--> D[/"SRT | ASS | TXT"/]
47 | ```
48 |
49 | ### Installation
50 |
51 | ```bash
52 | pip install rapid_videocr
53 | ```
54 |
55 | ### Usage
56 |
57 | > [!NOTE]
58 | >
59 | > The input image path of `rapid_videocr` must be the path of **RGBImages** or **TXTImages** output by **VideoSubFinder** software.
60 |
61 | ```bash
62 | rapid_videocr -i RGBImages
63 | ```
64 |
65 | or python script:
66 |
67 | ```python
68 | from rapid_videocr import RapidVideOCR, RapidVideOCRInput
69 |
70 | input_args = RapidVideOCRInput(is_batch_rec=False)
71 | extractor = RapidVideOCR(input_args)
72 |
73 | rgb_dir = "tests/test_files/RGBImages"
74 | save_dir = "outputs"
75 | save_name = "a"
76 |
77 | # outputs/a.srt outputs/a.ass outputs/a.txt
78 | extractor(rgb_dir, save_dir, save_name=save_name)
79 | ```
80 |
81 | ### Documentation
82 |
83 | Full documentation can be found on [docs](https://swhl.github.io/RapidVideOCR) in Chinese.
84 |
85 | ### Code Contributors
86 |
87 |
39 |
88 |
89 |
90 |
91 |
105 | 
106 |
107 |
108 | ### License
109 |
110 | This project is released under the [Apache 2.0 license](./LICENSE).
111 |
--------------------------------------------------------------------------------
/rapid_videocr/main.py:
--------------------------------------------------------------------------------
1 | # -*- encoding: utf-8 -*-
2 | # @Author: SWHL
3 | # @Contact: liekkaskono@163.com
4 | import argparse
5 | from pathlib import Path
6 | from typing import List, Union
7 |
8 | from .export import ExportStrategyFactory, OutputFormat
9 | from .ocr_processor import OCRProcessor
10 | from .utils.crop_by_project import CropByProject
11 | from .utils.logger import logger
12 | from .utils.typings import IMAGE_EXTENSIONS, RapidVideOCRInput
13 | from .utils.utils import mkdir
14 |
15 |
16 | class RapidVideOCR:
17 | def __init__(self, input_params: RapidVideOCRInput):
18 | logger.setLevel(input_params.log_level.upper())
19 |
20 | self.ocr_processor = OCRProcessor(
21 | input_params.ocr_params, input_params.batch_size
22 | )
23 |
24 | self.cropper = CropByProject()
25 |
26 | self.is_batch_rec = input_params.is_batch_rec
27 | self.out_format = input_params.out_format
28 |
29 | def __call__(
30 | self,
31 | vsf_dir: Union[str, Path],
32 | save_dir: Union[str, Path],
33 | save_name: str = "result",
34 | ) -> List[str]:
35 | vsf_dir = Path(vsf_dir)
36 | if not vsf_dir.exists():
37 | raise RapidVideOCRExeception(f"{vsf_dir} does not exist.")
38 |
39 | img_list = self.get_img_list(vsf_dir)
40 | srt_result, ass_result, txt_result = self.ocr_processor(
41 | img_list, self.is_batch_rec, self.is_txt_dir(vsf_dir)
42 | )
43 |
44 | self.export_file(Path(save_dir), save_name, srt_result, ass_result, txt_result)
45 | return txt_result
46 |
47 | def get_img_list(self, vsf_dir: Path) -> List[Path]:
48 | def get_sort_key(x: Path) -> int:
49 | return int("".join(str(x.stem).split("_")[:4]))
50 |
51 | img_list = []
52 | for v in vsf_dir.glob("*.*"):
53 | if not v.is_file():
54 | continue
55 |
56 | if v.suffix.lower() not in IMAGE_EXTENSIONS:
57 | continue
58 |
59 | img_list.append(v)
60 |
61 | if not img_list:
62 | raise RapidVideOCRExeception(f"{vsf_dir} does not have valid images")
63 |
64 | img_list = sorted(img_list, key=get_sort_key)
65 | return img_list
66 |
67 | @staticmethod
68 | def is_txt_dir(vsf_dir: Path) -> bool:
69 | return "TXTImages" in vsf_dir.name
70 |
71 | def export_file(
72 | self,
73 | save_dir: Path,
74 | save_name: str,
75 | srt_result: List[str],
76 | ass_result: List[str],
77 | txt_result: List[str],
78 | ):
79 | try:
80 | strategy = ExportStrategyFactory.create_strategy(self.out_format)
81 | mkdir(save_dir)
82 | strategy.export(save_dir, save_name, srt_result, ass_result, txt_result)
83 | logger.info(f"[OCR] Results saved to directory: {save_dir}")
84 | except ValueError as e:
85 | logger.error(f"Export failed: {e}")
86 | raise
87 |
88 | def print_console(self, txt_result: List):
89 | for v in txt_result:
90 | print(v.strip())
91 |
92 |
93 | class RapidVideOCRExeception(Exception):
94 | pass
95 |
96 |
97 | def main():
98 | parser = argparse.ArgumentParser()
99 | parser.add_argument(
100 | "-i",
101 | "--img_dir",
102 | type=str,
103 | required=True,
104 | help="The full path of RGBImages or TXTImages.",
105 | )
106 | parser.add_argument(
107 | "-s",
108 | "--save_dir",
109 | type=str,
110 | default="outputs",
111 | help='The path of saving the recognition result. Default is "outputs" under the current directory.',
112 | )
113 | parser.add_argument(
114 | "-f",
115 | "--file_name",
116 | type=str,
117 | default="result",
118 | help='The name of the resulting file name. Default is "result".',
119 | )
120 | parser.add_argument(
121 | "-o",
122 | "--out_format",
123 | type=str,
124 | default=OutputFormat.ALL.value,
125 | choices=[v.value for v in OutputFormat],
126 | help='Output file format. Default is "all".',
127 | )
128 | parser.add_argument(
129 | "--is_batch_rec",
130 | action="store_true",
131 | default=False,
132 | help="Which mode to run (concat recognition or single recognition). Default is False.",
133 | )
134 | parser.add_argument(
135 | "-b",
136 | "--batch_size",
137 | type=int,
138 | default=10,
139 | help="The batch of concating image nums in concat recognition mode. Default is 10.",
140 | )
141 | args = parser.parse_args()
142 |
143 | ocr_input_params = RapidVideOCRInput(
144 | is_batch_rec=args.is_batch_rec,
145 | batch_size=args.batch_size,
146 | out_format=args.out_format,
147 | )
148 | extractor = RapidVideOCR(ocr_input_params)
149 | extractor(args.img_dir, args.save_dir, args.file_name)
150 |
151 |
152 | if __name__ == "__main__":
153 | main()
154 |
--------------------------------------------------------------------------------
/docs/tutorial/intermediate.md:
--------------------------------------------------------------------------------
1 | ---
2 | comments: true
3 | hide:
4 | - toc
5 | ---
6 |
7 |
8 | ## 引言
9 |
10 | - 本篇文章旨在帮助不会python编程的小伙伴，快速使用RapidVideOCR视频硬字幕提取工具。
11 | - 可以运行的操作系统: `Windows | Mac | Linux`
12 |
13 | ## 配置环境
14 |
15 | ### 1. 安装VideoSubFinder软件（用于提取字幕关键帧）
16 |
17 | [[RapidVideOCR周边] VideoSubFinder提取字幕关键帧教程](https://blog.csdn.net/shiwanghualuo/article/details/129174857?spm=1001.2014.3001.5501)
18 |
19 | ### 2. 安装python软件（用于运行RapidVideOCR）
20 |
21 | !!! tip
22 |
23 | 声明：以下部分图像来自[终极保姆教程之安装python的教程_3.10.7版本](https://www.cnblogs.com/zyc-666/p/16689739.html)
24 |
25 | #### 1. 下载python安装包
26 |
27 | 打开python官网 →
106 |
30 | 
31 |
32 |
33 | #### 2. 找到自己想要的版本
34 |
35 | 以python 3.10.7 为例。如果网速下载较慢的话，可以加入RapidVideOCR QQ群（706807542），群文件获取。
36 |
37 |
31 |
38 | 
39 |
40 |
41 | #### 3. 下载完成之后，双击打开这个exe，即可开始准备安装
42 |
43 | 点击自定义安装，选择安装位置。同时，记得勾选最后一项`Add Python 3.10 to PATH`
44 |
45 |
39 |
46 | 
47 |
48 |
49 | #### 4. 直接点击下一步
50 |
51 |
47 |
52 | 
53 |
54 |
55 | #### 5. 勾选改路径
56 |
57 |
53 |
58 | 
59 |
60 |
61 | #### 6. 点击Install，等待安装完毕即可
62 |
63 |
59 |
64 | 
65 |
66 |
67 | #### 7. `Win + r` 输入`cmd`，回车，进入命令窗口
68 |
69 |
65 |
70 | 
71 |
72 |
73 | #### 8. 输入python，看是否出现类似下图样子，如出现，证明安装成功
74 |
75 |
71 |
76 | 
77 |
78 |
79 | #### 9. 添加`Scripts`目录到环境变量中
80 |
81 | 1. `Win + q` 输入“编辑” → 点击 **编辑系统环境变量**
82 |
77 |
83 | 
84 |
85 |
86 | 2. 打开 **环境变量** → **用户变量** → **Path** → **编辑**
87 |
84 |
88 | 
89 |
90 |
91 | 3. 新建Python安装目录下的Script目录路径，如下图所示，记得点击保存哈。
92 |
93 |
89 |
94 | 
95 |
96 |
97 | ### 3. 安装RapidVideOCR工具
98 |
99 | #### 1. `Win + r` 输入`cmd`，回车，进入命令窗口
100 |
101 |
95 |
102 | 
103 |
104 |
105 | #### 2. 安装`rapid_videocr`
106 |
107 | ```bash linenums="1"
108 | pip install rapid_videocr -i https://pypi.tuna.tsinghua.edu.cn/simple/
109 | ```
110 |
111 |
103 |
112 | 
113 |
114 |
115 | #### 3. 测试是否安装成功，输入`rapid_videocr -h`,如果出现类似下图输出，则说明安装成功
116 |
117 |
113 |
118 | 
119 |
120 |
121 | #### 4. 命令行使用
122 |
123 | `Win + r` 输入`cmd`，回车，进入命令行窗口
124 |
125 | ```bash linenums="1"
126 | rapid_videocr -i RGBImages -s result -m concat
127 | ```
128 |
129 | 其中`RGBImages`为VideoSubFinder软件生成，可以自定义，例如：`G:\ProgramFiles\_self\RapidVideOCR\test_files\RGBImages` 等等。
130 |
131 |
119 |
132 | 
133 |
134 |
135 | #### 5. 脚本使用
136 |
137 | 1. 在桌面上新建TXT文件，命名为`rapid_videocr.py`，注意后缀名改为`.py`。
138 | 2. 用记事本打开，将以下代码拷贝到`rapid_videocr.py`里面
139 |
140 | ```python linenums="1"
141 | from rapid_videocr import RapidVideOCR, RapidVideOCRInput
142 |
143 | # RapidVideOCRInput有两个初始化参数
144 | # is_concat_rec: 是否用单张图识别，默认是False，也就是默认用单图识别
145 | # concat_batch: 叠图识别的图像张数，默认10，可自行调节
146 | # out_format: 输出格式选择，[srt, ass, txt, all], 默认是 all
147 | # is_print_console: 是否打印结果，[0, 1], 默认是0，不打印
148 | input_args = RapidVideOCRInput(
149 | is_batch_rec=False, ocr_params={"Global.with_paddle": True}
150 | )
151 | extractor = RapidVideOCR(input_args)
152 |
153 | rgb_dir = "tests/test_files/RGBImages"
154 | save_dir = "outputs"
155 | save_name = "a"
156 |
157 | # outputs/a.srt outputs/a.ass outputs/a.t
158 | extractor(rgb_dir, save_dir, save_name=save_name)
159 | ```
160 |
161 | 3. 更改`rgb_dir` 后面的目录为VideoSubFinder生成的`RGBImages`目录路径。
162 |
163 | === "Windows下路径写法"
164 |
165 | ```python linenums="1"
166 | rgb_dir = r'G:\ProgramFiles\_self\RapidVideOCR\test_files\RGBImages'
167 | ```
168 |
169 | === "Linux/Mac下路径写法"
170 |
171 | ```python linenums="1"
172 | rgb_dir = 'test_files/TXTImages'
173 | ```
174 |
175 | 4. `Win + r` 打开终端输入以下代码，回车执行即可。
176 |
177 | ```bash linenums="1"
178 | cd Desktop
179 | python rapid_videocr.py
180 | ```
181 |
182 |
133 |
183 | 
184 |
185 |
--------------------------------------------------------------------------------
/docs/changelog.en.md:
--------------------------------------------------------------------------------
1 | ---
2 | comments: true
3 | hide:
4 | - toc
5 | ---
6 |
7 | ### 📣 Subsequent update logs will be moved to [Release](https://github.com/SWHL/RapidVideOCR/releases), and will no longer be updated here
8 |
9 | #### 🚩2023-10-08 v2.2.8 update
10 |
11 | - Adapt the relevant parameters of `rapidocr_onnxruntime`, which can be passed in through the RapidVideOCR class, so as to more flexibly specify models of different languages.
12 |
13 | #### ♦ 2023-08-05 v2.2.4 update
14 |
15 | - Fix the index error in batch recognition mode.
16 |
17 | - Add a logging module to facilitate the use of the desktop version, quickly record problems, and facilitate feedback.
18 |
19 | #### 🛶2023-07-19 v2.2.3 update
20 |
21 | - Added adaptation to VSF parameters. When in command line mode and during class initialization, you can specify the same-name parameters of the VSF command. For detailed usage, please refer to [link](https://github.com/SWHL/RapidVideOCR/wiki/RapidVideOCR%E9%AB%98%E7%BA%A7%E6%95%99%E7%A8%8B%EF%BC%88%E6%9C%89python%E5%9F%BA%E7%A1%80%E7%9A%84%E5%B0%8F%E4%BC%99%E4%BC%B4%EF%BC%89)
22 |
23 | #### 🤓2023-07-08 v2.2.2 update
24 |
25 | - Fixed the problem that Chinese paths cannot be read during batch recognition
26 | - Fixed the problem of skipping in SRT when missing axes. Currently, when an axis fails to be recognized, a position will be vacated for easy proofreading.
27 | - Keep the intermediate results of VSF recognition
28 |
29 | #### 🐲2023-06-22 v2.2.0 update
30 |
31 | - This version is compatible with `v2.1.x`, which means that the previous usage is still possible.
32 | - Integrate VSF's CLI into the library, just specify the full path of `VideoSubFinderWXW.exe`.
33 | - Added batch recognition function, specify the video directory, and automatically extract all video subtitles in the directory
34 | - For usage examples, see: [demo.py](https://github.com/SWHL/RapidVideOCR/blob/main/demo.py)
35 |
36 | #### 😀2023-05-12 v2.1.7 update
37 |
38 | - Optimized code
39 | - Added `save_name` parameter, which can flexibly specify the name of the saved `srt | txt` file, the default is `result`
40 |
41 | #### 🐱2023-03-27 v2.1.6 update
42 |
43 | - Fixed the timeline misalignment problem, see [issue 23](https://github.com/SWHL/RapidVideOCR/issues/23) for details
44 |
45 | #### 👽2023-03-23 v2.1.5 update
46 |
47 | - Added control parameter `is_print_console` for printing to the screen
48 | - Adjust the `out_format` parameter position to the initialization class
49 |
50 | #### 😀2023-03-14 v2.1.3 update
51 |
52 | - Fix the error when passing in the `TXTImages` directory and identifying duplicate characters
53 |
54 | #### 😜2023-03-12 v2.1.2 update
55 |
56 | - Fix index error, [#22](https://github.com/SWHL/RapidVideOCR/issues/22)
57 |
58 | #### 🎢2023-03-11 v2.1.1 update
59 |
60 | - Fix the difference between single image recognition and previous versions
61 |
62 | - The default recognition mode is changed to single image recognition. Whether to use duplicate image recognition is up to you
63 |
64 | #### 🥇2023-03-10 v2.1.0 update
65 |
66 | - Added duplicate character recognition function, faster, and the default is duplicate character recognition function
67 |
68 | #### 🎈2023-03-02 v2.0.5~7 update
69 |
70 | - Fix format errors in generated srt files, [#19](https://github.com/SWHL/RapidVideOCR/issues/19)
71 |
72 | #### 🎫2023-02-17 v2.0.4 update
73 |
74 | - Optimized passing in the `TXTImages` directory. Compared with passing in `RGBImages`, it will be faster and more accurate. It is recommended to pass in the `TXTImages` directory
75 |
76 | #### 💎2023-02-17 v2.0.2 update
77 |
78 | - Fix the problem of missing spaces in peer subtitle recognition
79 |
80 | #### 🎈2023-01-29 v1.1.10 update
81 |
82 | - Fix the error of empty index when converting frame index to timestamp
83 |
84 | #### 🧨2023-01-28 v1.1.9 update
85 |
86 | - Fix the problem of timeline misalignment
87 |
88 | #### 👊 2023-01-15 v1.1.4 update
89 |
90 | - Add the option of outputting txt format. Currently, the v1.1.4 version outputs srt and txt formats by default
91 |
92 | - Add the option to adjust the box size of subtitle selection according to the screen size of the running program
93 |
94 | #### 🌈2023-01-10 v1.0.3 update
95 |
96 | - Replace decord with OpenCV because decord has a memory leak when processing MP4. For details, see: [#208](https://github.com/dmlc/decord/issues/208)
97 |
98 | #### 🎄2022-12-04 update
99 |
100 | - Added interactive subtitle positioning function, which is enabled by default and is more user-friendly. For details, please refer to the GIF below. Thanks to @[Johndirr](https://github.com/Johndirr) for the suggestion.
101 | - Optimize the code structure, put RapidOCR related models and configuration files in the `rapidocr` directory
102 | - The configuration files of `rapidvideocr` are also placed in the corresponding directory
103 |
104 | #### 🌼2022-05-08 update
105 |
106 | - Add interactive determination of the threshold of the binary subtitle image, which only supports Windows and can be used by `is_select_threshold = True`
107 | - Optimize the code
108 |
109 | #### 🎉2022-05-03 update
110 |
111 | - Add GPU support, for specific configuration tutorials, see: [onnxruntime-gpu version inference configuration](https://github.com/RapidAI/RapidOCR/blob/main/python/onnxruntime_infer/README.md#onnxruntime-gpu%E7%89%88%E6%8E%A8%E7%90%86%E9%85%8D%E7%BD%AE)
112 | - Added support for Japanese, which can support more languages. For details, see: [Supported Language List](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_ch/multi_languages.md#%E8%AF%AD%E7%A7%8D%E7%BC%A9%E5%86%99)
113 |
114 | #### 💡2022-05-01 update
115 |
116 | - Added speech module part is located in the branch `asr_module`
117 |
118 | - Added speech recognition module. Since the decoding part of this module can only run on Linux and Mac, if you want to use this module, please use Linux and Mac.
119 | - Currently, the speech recognition code comes from the [RapidASR/python](https://github.com/RapidAI/RapidASR/tree/main/python/base_paddlespeech) part. The model comes from [PaddleSpeech](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/aishell/asr0)
120 | - After a simple test, the speech recognition module is not very accurate. -_-!
121 |
122 | #### 2022-03-09 update
123 |
124 | - Added [FAQ module](./faq.md) to help everyone skip common small problems
125 |
126 | #### 2021-12-14 update
127 |
128 | - [x] Background removal is not effective, resulting in the loss of some frames
129 |
130 | - Tried to use the image segmentation method. After testing, the inference speed under CPU is too slow, so it is temporarily abandoned
131 |
132 | - Currently using a fixed binary threshold
133 |
134 | - [x] (2021-12-14) Improve the corresponding English document
135 |
136 | - [x] (2021-12-14) Add running time benchmark
137 |
138 | - [x] Add specific parameter description
139 |
140 | - [x] Make a project logo
141 |
142 | - [ ] More tests
143 |
--------------------------------------------------------------------------------
/docs/tutorial/intermediate.en.md:
--------------------------------------------------------------------------------
1 | ---
2 | comments: true
3 | hide:
4 | - toc
5 | ---
6 |
7 |
8 | ## Introduction
9 |
10 | This article is aimed towards helping beginner Python programmers to quickly start using the RapidVideOCR subtitle extraction tool
11 |
12 | Supported operating systems: `Windows | Mac | Linux`
13 |
14 | ## Environment Configuration
15 |
16 | ### 1. Install VideoSubFinder (used for extracting subtitle keyframes)
17 |
18 | [[RapidVideOCR extras] VideoSubFinder subtitle keyframe extraction tutorial](https://blog.csdn.net/shiwanghualuo/article/details/129174857?spm=1001.2014.3001.5501)
19 |
20 | ### 2. Install python（used for running RapidVideOCR）
21 |
22 | !!! tip
23 |
24 | Disclaimer: Some of the following images are from [终极保姆教's tutorial on installing Python_3.10.7](https://www.cnblogs.com/zyc-666/p/16689739.html)
25 |
26 | #### 1. Download the Python installer
27 |
28 | Open the official Python website →
184 |
31 |
32 |
33 |
34 | #### 2. Find the version you want
35 |
36 | Take Python 3.10.7 for example, if the download speed is too slow you can join the RapidVideOCR QQ group: (706807542) to download the files.
37 |
38 |
32 |
39 |
40 |
41 |
42 | #### 3. After the download is complete, double-click to open the exe and start the installation
43 |
44 | Click Customize installation and select the installation location. Also, remember to check the last item `Add Python 3.10 to PATH`
45 |
46 |
40 |
47 |
48 |
49 |
50 | #### 4. Click Next
51 |
52 |
48 |
53 |
54 |
55 |
56 | #### 5. Check the installation path
57 |
58 |
54 |
59 |
60 |
61 |
62 | #### 6. Click install, and wait for the installation to finish
63 |
64 |
60 |
65 |
66 |
67 |
68 | #### 7. Press `Win + r` input `cmd`, and press Enter to open the command prompt
69 |
70 |
66 |
71 |
72 |
73 |
74 | #### 8. Enter python and see if an output something similar to the following image appears. If so, then the installation was successful
75 |
76 |
72 |
77 |
78 |
79 |
80 | #### 9. Add the `Scripts` directory to the environment variables
81 |
82 | 1. Press `Win + q` enter `Edit the system`→ Click **Edit the system environment variables**
83 |
78 |
84 |
85 |
86 |
87 | 2. Click **Environment Variables** → **User variables** → **Path** → **Edit**
88 |
85 |
89 |
90 |
91 |
92 | 3. Create a new entry for the Script directory under the Python installation directory, as shown in the image below, and remember to click save.
93 |
94 |
90 |
95 |
96 |
97 |
98 | ### 3. Install RapidVideOCR
99 |
100 | #### 1. Press `Win + r` input `cmd`, and press Enter to open the command prompt
101 |
102 |
96 |
103 |
104 |
105 |
106 | #### 2. Install `rapid_videocr`
107 |
108 | ```bash linenums="1"
109 | pip install rapid_videocr -i https://pypi.tuna.tsinghua.edu.cn/simple/
110 | ```
111 |
112 |
104 |
113 |
114 |
115 |
116 | #### 3. To test whether the installation was successful, enter `rapid_videocr -h`
117 |
118 |
114 |
119 |
120 |
121 |
122 | #### 4. Command line usage
123 |
124 | Press `Win + r` input `cmd`, and press Enter to open the command prompt
125 |
126 | ```bash linenums="1"
127 | rapid_videocr -i RGBImages -s result -m concat
128 | ```
129 |
130 | `RGBImages` is generated by VideoSubFinder and its output path can be customized, for example: `G:\ProgramFiles\_self\RapidVideOCR\test_files\RGBImages` and so on.
131 |
132 |
120 |
133 |
134 |
135 |
136 | #### 5. Script usage
137 |
138 | 1. Create a new TXT file on the desktop and name it `rapid_videocr.py`. Note that the file extension is changed to `*.py`.
139 | 2. Open `rapid_videocr.py` with Notepad and copy the following code into it
140 |
141 | ```python linenums="1"
142 | from rapid_videocr import RapidVideOCR, RapidVideOCRInput
143 |
144 | # RapidVideOCRInput has two initialization parameters
145 | # is_concat_rec: Use a single image for recognition or not. The default is False, which means that a single image is used for recognition by default.
146 | # concat_batch: The number of images to be used in overlay is 10 by default and can be adjusted
147 | # out_format: Output format selection, [srt, ass, txt, all], the default is all
148 | # is_print_console: Whether to print the result, [0, 1], the default is 0 for not printing
149 | input_args = RapidVideOCRInput(
150 | is_batch_rec=False, ocr_params={"Global.with_paddle": True}
151 | )
152 | extractor = RapidVideOCR(input_args)
153 |
154 | rgb_dir = "tests/test_files/RGBImages"
155 | save_dir = "outputs"
156 | save_name = "a"
157 |
158 | # outputs/a.srt outputs/a.ass outputs/a.t
159 | extractor(rgb_dir, save_dir, save_name=save_name)
160 | ```
161 |
162 | 3. Change `rgb_dir` to the path to the `RGBImages` directory generated by VideoSubFinder。
163 |
164 | === "Windows path style"
165 |
166 | ```python linenums="1"
167 | rgb_dir = r'G:\ProgramFiles\_self\RapidVideOCR\test_files\RGBImages'
168 | ```
169 |
170 | === "Linux/Mac path style"
171 |
172 | ```python linenums="1"
173 | rgb_dir = 'test_files/TXTImages'
174 | ```
175 |
176 | 4. Press `Win + r` and open the command prompt, and run the following commands
177 |
178 | ```bash linenums="1"
179 | cd Desktop
180 | python rapid_videocr.py
181 | ```
182 |
183 |
134 |
184 |
185 |
186 |
--------------------------------------------------------------------------------
/docs/tutorial/senior.md:
--------------------------------------------------------------------------------
1 | ---
2 | comments: true
3 | hide:
4 | - toc
5 | ---
6 |
7 | ### 1. 安装使用VideoSubFinder软件
8 |
9 | 下载地址：Windows & Linux ([videosubfinder官网](https://sourceforge.net/projects/videosubfinder/) / QQ群（706807542）共享文件) | [Mac版](https://github.com/eritpchy/videosubfinder-cli)
10 |
11 | 使用教程：[VideoSubFinder提取字幕关键帧教程](https://juejin.cn/post/7203362527082053691)
12 |
13 | 最终生成的`RGBImages`和`TXTImages`目录一般会在软件安装目录下
14 |
15 | ✧ 推荐用`RGBImages`目录中图像（感谢小伙伴[dyphire](https://github.com/dyphire)在[#21](https://github.com/SWHL/RapidVideOCR/issues/21)的反馈）
16 |
17 | ### 2. 安装rapid_videocr
18 |
19 | ```bash linenums="1"
20 | pip install rapid_videocr
21 | ```
22 |
23 | ### 3. Python使用
24 |
25 | === "Only OCR"
26 |
27 | ```python linenums="1"
28 | from rapid_videocr import RapidVideOCR
29 |
30 | # RapidVideOCRInput有两个初始化参数
31 | # is_concat_rec: 是否用单张图识别，默认是False，也就是默认用单图识别
32 | # concat_batch: 叠图识别的图像张数，默认10，可自行调节
33 | # out_format: 输出格式选择，[srt, ass, txt, all], 默认是 all
34 | # is_print_console: 是否打印结果，[0, 1], 默认是0，不打印
35 | input_args = RapidVideOCRInput(
36 | is_batch_rec=False, ocr_params={"Global.with_paddle": True}
37 | )
38 | extractor = RapidVideOCR(input_args)
39 |
40 | rgb_dir = "tests/test_files/RGBImages"
41 | save_dir = "outputs"
42 | save_name = "a"
43 |
44 | # outputs/a.srt outputs/a.ass outputs/a.t
45 | extractor(rgb_dir, save_dir, save_name=save_name)
46 | ```
47 |
48 | === "Extract + OCR"
49 |
50 | ```python linenums="1"
51 | from rapid_videocr import RapidVideoSubFinderOCR
52 |
53 | vsf_exe = r"G:\ProgramFiles\VideoSubFinder_6.10_x64\Release_x64\VideoSubFinderWXW.exe"
54 | extractor = RapidVideoSubFinderOCR(vsf_exe_path=vsf_exe, is_concat_rec=True)
55 |
56 | # video_path can be directory path or video full path.
57 | video_path = 'test_files/tiny/2.mp4'
58 | save_dir = 'outputs'
59 | extractor(video_path, save_dir)
60 | ```
61 |
62 | ### 4. 命令行使用
63 |
64 | === "Only OCR"
65 |
66 | ```bash linenums="1"
67 | rapid_videocr -i RGBImages
68 | ```
69 |
70 | === "Extract + OCR"
71 |
72 | ```bash linenums="1"
73 | rapid_videocr -vsf G:\ProgramFiles\VideoSubFinder_6.10_x64\Release_x64\VideoSubFinderWXW.exe -video_dir G:\ProgramFiles\RapidVideOCR\test_files\tiny
74 | ```
75 |
76 |
185 |
77 |
78 |
164 |
165 | ### 5. 查看结果
166 |
167 | !!! info
168 |
169 | "如果想要让视频播放软件自动挂载srt文件或ass文件，需要更改srt或ass文件名字为视频文件名字，且放到同一目录下，亦或者手动指定加载。
170 |
171 | 前往`save_dir`目录下即可查看结果。
172 |
--------------------------------------------------------------------------------
/docs/tutorial/senior.en.md:
--------------------------------------------------------------------------------
1 | ---
2 | comments: true
3 | hide:
4 | - toc
5 | ---
6 |
7 | ### 1. Install and use VideoSubFinder
8 |
9 | - Download link: Windows & Linux ([videosubfinder](https://sourceforge.net/projects/videosubfinder/) / QQ group (706807542) shared files) | [Mac version](https://github.com/eritpchy/videosubfinder-cli)
10 | - Tutorial: [Tutorial for extracting subtitle keyframes with VideoSubFinder](https://juejin.cn/post/7203362527082053691)
11 | - The final `RGBImages` and `TXTImages` are usually in the installation directory for VideoSubFinder
12 | - ✧ It is recommended to use `RGBImages` (thanks to [dyphire](https://github.com/dyphire) for the feedback in [#21](https://github.com/SWHL/RapidVideOCR/issues/21))
13 |
14 | ### 2. Install rapid_videocr
15 |
16 | ```bash linenums="1"
17 | pip install rapid_videocr
18 | ```
19 |
20 | ### 3. Python usage
21 |
22 | {{< tabs tabTotal="2">}}
23 | {{% tab tabName="Only OCR" %}}
24 |
25 | ```python linenums="1"
26 | from rapid_videocr import RapidVideOCR
27 |
28 | # RapidVideOCRInput has two initialization parameters
29 | # is_concat_rec: Use a single image for recognition or not. The default is False, which means that a single image is used for recognition by default.
30 | # concat_batch: The number of images to be used in overlay is 10 by default and can be adjusted
31 | # out_format: Output format selection, [srt, ass, txt, all], the default is all
32 | # is_print_console: Whether to print the result, [0, 1], the default is 0 for not printing
33 | ocr_input_params = RapidVideOCRInput(
34 | is_batch_rec=False, ocr_params={"Global.with_paddle": True}
35 | )
36 | extractor = RapidVideOCR(ocr_input_params)
37 |
38 | rgb_dir = "tests/test_files/RGBImages"
39 | save_dir = "outputs"
40 | save_name = "a"
41 |
42 | # outputs/a.srt outputs/a.ass outputs/a.t
43 | extractor(rgb_dir, save_dir, save_name=save_name)
44 | ```
45 |
46 | {{% /tab %}}
47 | {{% tab tabName="Extract + OCR" %}}
48 |
49 | ```python linenums="1"
50 | from rapid_videocr import RapidVideoSubFinderOCR
51 |
52 | vsf_exe = r"G:\ProgramFiles\VideoSubFinder_6.10_x64\Release_x64\VideoSubFinderWXW.exe"
53 | extractor = RapidVideoSubFinderOCR(vsf_exe_path=vsf_exe, is_concat_rec=True)
54 |
55 | # video_path can be directory path or video full path.
56 | video_path = 'test_files/tiny/2.mp4'
57 | save_dir = 'outputs'
58 | extractor(video_path, save_dir)
59 | ```
60 |
61 | {{% /tab %}}
62 | {{< /tabs >}}
63 |
64 | ### 4. Command line usage
65 |
66 | {{< tabs tabTotal="2">}}
67 | {{% tab tabName="Only OCR" %}}
68 |
69 | ```bash linenums="1"
70 | rapid_videocr -i RGBImages
71 | ```
72 |
73 | {{% /tab %}}
74 | {{% tab tabName="Extract + OCR" %}}
75 |
76 | ```bash linenums="1"
77 | rapid_videocr -vsf G:\ProgramFiles\VideoSubFinder_6.10_x64\Release_x64\VideoSubFinderWXW.exe -video_dir G:\ProgramFiles\RapidVideOCR\test_files\tiny
78 | ```
79 |
80 | {{% /tab %}}
81 | {{< /tabs >}}
82 |
83 | Parameter details:
84 | 详细参数 79 | 80 | ```bash linenums="1" 81 | $ rapid_videocr -h 82 | usage: rapid_videocr [-h] [-video_dir VIDEO_DIR] [-i IMG_DIR] [-s SAVE_DIR] 83 | [-o {srt,ass,txt,all}] [--is_concat_rec] [-b CONCAT_BATCH] [-p] 84 | [-vsf VSF_EXE_PATH] [-c] [-r] [-ccti] [-ces CREATE_EMPTY_SUB] 85 | [-cscti CREATE_SUB_FROM_CLEARED_TXT_IMAGES] 86 | [-cstxt CREATE_SUB_FROM_TXT_RESULTS] [-ovocv] [-ovffmpeg] [-uc] 87 | [--start_time START_TIME] [--end_time END_TIME] 88 | [-te TOP_VIDEO_IMAGE_PERCENT_END] 89 | [-be BOTTOM_VIDEO_IMAGE_PERCENT_END] 90 | [-le LEFT_VIDEO_IMAGE_PERCENT_END] 91 | [-re RIGHT_VIDEO_IMAGE_PERCENT_END] [-gs GENERAL_SETTINGS] 92 | [-nthr NUM_THREADS] [-nocrthr NUM_OCR_THREADS] 93 | 94 | optional arguments: 95 | -h, --help show this help message and exit 96 | 97 | VideOCRParameters: 98 | -video_dir VIDEO_DIR, --video_dir VIDEO_DIR 99 | The full path of video or the path of video directory. 100 | -i IMG_DIR, --img_dir IMG_DIR 101 | The full path of RGBImages or TXTImages. 102 | -s SAVE_DIR, --save_dir SAVE_DIR 103 | The path of saving the recognition result. Default is 104 | "outputs" under the current directory. 105 | -o {srt,ass,txt,all}, --out_format {srt,ass,txt,all} 106 | Output file format. Default is "all". 107 | --is_concat_rec Which mode to run (concat recognition or single 108 | recognition). Default is False. 109 | -b CONCAT_BATCH, --concat_batch CONCAT_BATCH 110 | The batch of concating image nums in concat 111 | recognition mode. Default is 10. 112 | -p, --print_console Whether to print the subtitle results to console. -p 113 | means to print. 114 | 115 | VSFParameters: 116 | -vsf VSF_EXE_PATH, --vsf_exe_path VSF_EXE_PATH 117 | The full path of VideoSubFinderWXW.exe. 118 | -c, --clear_dirs Clear Folders (remove all images), performed before 119 | any other steps. Default is True 120 | -r, --run_search Run Search (find frames with hardcoded text (hardsub) 121 | on video) Default is True 122 | -ccti, --create_cleared_text_images 123 | Create Cleared Text Images. Default is True 124 | -ces CREATE_EMPTY_SUB, --create_empty_sub CREATE_EMPTY_SUB 125 | Create Empty Sub With Provided Output File Name (*.ass 126 | or *.srt) 127 | -cscti CREATE_SUB_FROM_CLEARED_TXT_IMAGES, --create_sub_from_cleared_txt_images CREATE_SUB_FROM_CLEARED_TXT_IMAGES 128 | Create Sub From Cleared TXT Images With Provided 129 | Output File Name (*.ass or *.srt) 130 | -cstxt CREATE_SUB_FROM_TXT_RESULTS, --create_sub_from_txt_results CREATE_SUB_FROM_TXT_RESULTS 131 | Create Sub From TXT Results With Provided Output File 132 | Name (*.ass or *.srt) 133 | -ovocv, --open_video_opencv 134 | open video by OpenCV (default). Default is True 135 | -ovffmpeg, --open_video_ffmpeg 136 | open video by FFMPEG 137 | -uc, --use_cuda use cuda 138 | --start_time START_TIME 139 | start time, default = 0:00:00:000 (in format 140 | hour:min:sec:milisec) 141 | --end_time END_TIME end time, default = video length 142 | -te TOP_VIDEO_IMAGE_PERCENT_END, --top_video_image_percent_end TOP_VIDEO_IMAGE_PERCENT_END 143 | top video image percent offset from image bottom, can 144 | be in range [0.0,1.0], default = 1.0 145 | -be BOTTOM_VIDEO_IMAGE_PERCENT_END, --bottom_video_image_percent_end BOTTOM_VIDEO_IMAGE_PERCENT_END 146 | bottom video image percent offset from image bottom, 147 | can be in range [0.0,1.0], default = 0.0 148 | -le LEFT_VIDEO_IMAGE_PERCENT_END, --left_video_image_percent_end LEFT_VIDEO_IMAGE_PERCENT_END 149 | left video image percent end, can be in range 150 | [0.0,1.0], default = 0.0 151 | -re RIGHT_VIDEO_IMAGE_PERCENT_END, --right_video_image_percent_end RIGHT_VIDEO_IMAGE_PERCENT_END 152 | right video image percent end, can be in range 153 | [0.0,1.0], default = 1.0 154 | -gs GENERAL_SETTINGS, --general_settings GENERAL_SETTINGS 155 | general settings (path to general settings *.cfg file, 156 | default = settings/general.cfg) 157 | -nthr NUM_THREADS, --num_threads NUM_THREADS 158 | number of threads used for Run Search 159 | -nocrthr NUM_OCR_THREADS, --num_ocr_threads NUM_OCR_THREADS 160 | number of threads used for Create Cleared TXT Images 161 | ``` 162 | 163 |
85 |
86 | ```bash linenums="1"
87 | $ rapid_videocr -h
88 | usage: rapid_videocr [-h] [-video_dir VIDEO_DIR] [-i IMG_DIR] [-s SAVE_DIR]
89 | [-o {srt,ass,txt,all}] [--is_concat_rec] [-b CONCAT_BATCH] [-p]
90 | [-vsf VSF_EXE_PATH] [-c] [-r] [-ccti] [-ces CREATE_EMPTY_SUB]
91 | [-cscti CREATE_SUB_FROM_CLEARED_TXT_IMAGES]
92 | [-cstxt CREATE_SUB_FROM_TXT_RESULTS] [-ovocv] [-ovffmpeg] [-uc]
93 | [--start_time START_TIME] [--end_time END_TIME]
94 | [-te TOP_VIDEO_IMAGE_PERCENT_END]
95 | [-be BOTTOM_VIDEO_IMAGE_PERCENT_END]
96 | [-le LEFT_VIDEO_IMAGE_PERCENT_END]
97 | [-re RIGHT_VIDEO_IMAGE_PERCENT_END] [-gs GENERAL_SETTINGS]
98 | [-nthr NUM_THREADS] [-nocrthr NUM_OCR_THREADS]
99 |
100 | optional arguments:
101 | -h, --help show this help message and exit
102 |
103 | VideOCRParameters:
104 | -video_dir VIDEO_DIR, --video_dir VIDEO_DIR
105 | The full path of video or the path of video directory.
106 | -i IMG_DIR, --img_dir IMG_DIR
107 | The full path of RGBImages or TXTImages.
108 | -s SAVE_DIR, --save_dir SAVE_DIR
109 | The path of saving the recognition result. Default is
110 | "outputs" under the current directory.
111 | -o {srt,ass,txt,all}, --out_format {srt,ass,txt,all}
112 | Output file format. Default is "all".
113 | --is_concat_rec Which mode to run (concat recognition or single
114 | recognition). Default is False.
115 | -b CONCAT_BATCH, --concat_batch CONCAT_BATCH
116 | The batch of concating image nums in concat
117 | recognition mode. Default is 10.
118 | -p, --print_console Whether to print the subtitle results to console. -p
119 | means to print.
120 |
121 | VSFParameters:
122 | -vsf VSF_EXE_PATH, --vsf_exe_path VSF_EXE_PATH
123 | The full path of VideoSubFinderWXW.exe.
124 | -c, --clear_dirs Clear Folders (remove all images), performed before
125 | any other steps. Default is True
126 | -r, --run_search Run Search (find frames with hardcoded text (hardsub)
127 | on video) Default is True
128 | -ccti, --create_cleared_text_images
129 | Create Cleared Text Images. Default is True
130 | -ces CREATE_EMPTY_SUB, --create_empty_sub CREATE_EMPTY_SUB
131 | Create Empty Sub With Provided Output File Name (*.ass
132 | or *.srt)
133 | -cscti CREATE_SUB_FROM_CLEARED_TXT_IMAGES, --create_sub_from_cleared_txt_images CREATE_SUB_FROM_CLEARED_TXT_IMAGES
134 | Create Sub From Cleared TXT Images With Provided
135 | Output File Name (*.ass or *.srt)
136 | -cstxt CREATE_SUB_FROM_TXT_RESULTS, --create_sub_from_txt_results CREATE_SUB_FROM_TXT_RESULTS
137 | Create Sub From TXT Results With Provided Output File
138 | Name (*.ass or *.srt)
139 | -ovocv, --open_video_opencv
140 | open video by OpenCV (default). Default is True
141 | -ovffmpeg, --open_video_ffmpeg
142 | open video by FFMPEG
143 | -uc, --use_cuda use cuda
144 | --start_time START_TIME
145 | start time, default = 0:00:00:000 (in format
146 | hour:min:sec:milisec)
147 | --end_time END_TIME end time, default = video length
148 | -te TOP_VIDEO_IMAGE_PERCENT_END, --top_video_image_percent_end TOP_VIDEO_IMAGE_PERCENT_END
149 | top video image percent offset from image bottom, can
150 | be in range [0.0,1.0], default = 1.0
151 | -be BOTTOM_VIDEO_IMAGE_PERCENT_END, --bottom_video_image_percent_end BOTTOM_VIDEO_IMAGE_PERCENT_END
152 | bottom video image percent offset from image bottom,
153 | can be in range [0.0,1.0], default = 0.0
154 | -le LEFT_VIDEO_IMAGE_PERCENT_END, --left_video_image_percent_end LEFT_VIDEO_IMAGE_PERCENT_END
155 | left video image percent end, can be in range
156 | [0.0,1.0], default = 0.0
157 | -re RIGHT_VIDEO_IMAGE_PERCENT_END, --right_video_image_percent_end RIGHT_VIDEO_IMAGE_PERCENT_END
158 | right video image percent end, can be in range
159 | [0.0,1.0], default = 1.0
160 | -gs GENERAL_SETTINGS, --general_settings GENERAL_SETTINGS
161 | general settings (path to general settings *.cfg file,
162 | default = settings/general.cfg)
163 | -nthr NUM_THREADS, --num_threads NUM_THREADS
164 | number of threads used for Run Search
165 | -nocrthr NUM_OCR_THREADS, --num_ocr_threads NUM_OCR_THREADS
166 | number of threads used for Create Cleared TXT Images
167 | ```
168 |
169 |
170 |
171 | ### 5. View results
172 |
173 | Go to the `save_dir` directory to view the results.
174 |
175 | {{< alert context="info" text="If you want the video playback software to automatically mount the srt file or ass file, you need to change the srt or ass filename to be the same as the video file and put it in the same directory, or manually specify it." />}}
176 |
--------------------------------------------------------------------------------
/rapid_videocr/vsf_ocr_cli.py:
--------------------------------------------------------------------------------
1 | # -*- encoding: utf-8 -*-
2 | # @Author: SWHL
3 | # @Contact: liekkaskono@163.com
4 | import argparse
5 | from pathlib import Path
6 |
7 | from .main import OutputFormat, RapidVideOCR, RapidVideOCRInput
8 | from .utils.logger import logger
9 | from .utils.typings import VideoFormat
10 | from .utils.utils import float_range
11 | from .vsf_cli import VideoSubFinder, VideoSubFinderInput
12 |
13 |
14 | class RapidVideoSubFinderOCR:
15 | def __init__(
16 | self,
17 | vsf_input_params: VideoSubFinderInput,
18 | ocr_input_params: RapidVideOCRInput,
19 | ):
20 | self.vsf = VideoSubFinder(vsf_input_params)
21 | self.video_ocr = RapidVideOCR(ocr_input_params)
22 | self.video_formats = [VideoFormat[v].value for v in VideoFormat.__members__]
23 |
24 | def __call__(self, video_path: str, output_dir: str = "outputs"):
25 | if Path(video_path).is_dir():
26 | video_list = Path(video_path).rglob("*.*")
27 | video_list = [
28 | v for v in video_list if v.suffix.lower() in self.video_formats
29 | ]
30 | else:
31 | video_list = [video_path]
32 |
33 | logger.info(
34 | "Extracting subtitle images with VideoSubFinder (takes quite a long time) ..."
35 | )
36 | video_num = len(video_list)
37 | for i, one_video in enumerate(video_list):
38 | logger.info(
39 | f"[{i + 1}{video_num}] Starting to extract {one_video} key frame"
40 | )
41 |
42 | save_name = Path(one_video).stem
43 | save_dir = Path(output_dir) / save_name
44 | save_vsf_dir = save_dir / "VSF_Results"
45 |
46 | try:
47 | self.vsf(str(one_video), str(save_vsf_dir))
48 | except Exception as e:
49 | logger.error(f"Extract {one_video} error, {e}, skip")
50 | continue
51 |
52 | logger.info(f"{i + 1}/{video_num}] Starting to run {one_video} ocr")
53 |
54 | rgb_dir = Path(save_vsf_dir) / "RGBImages"
55 | if not list(rgb_dir.iterdir()):
56 | logger.warning(f"Extracting frames from {one_video} is 0, skip")
57 | continue
58 | self.video_ocr(rgb_dir, save_dir, save_name=save_name)
59 |
60 |
61 | def main():
62 | parser = argparse.ArgumentParser()
63 |
64 | videocr_param_group = parser.add_argument_group(title="VideOCRParameters")
65 | videocr_param_group.add_argument(
66 | "-video_dir",
67 | "--video_dir",
68 | type=str,
69 | default=None,
70 | help="The full path of video or the path of video directory.",
71 | )
72 | videocr_param_group.add_argument(
73 | "-i",
74 | "--img_dir",
75 | type=str,
76 | default=None,
77 | help="The full path of RGBImages or TXTImages.",
78 | )
79 | videocr_param_group.add_argument(
80 | "-s",
81 | "--save_dir",
82 | type=str,
83 | default="outputs",
84 | help='The path of saving the recognition result. Default is "outputs" under the current directory.',
85 | )
86 | videocr_param_group.add_argument(
87 | "-o",
88 | "--out_format",
89 | type=str,
90 | default=OutputFormat.ALL.value,
91 | choices=[OutputFormat[v].value for v in OutputFormat.__members__],
92 | help='Output file format. Default is "all".',
93 | )
94 | videocr_param_group.add_argument(
95 | "--is_batch_rec",
96 | action="store_true",
97 | default=False,
98 | help="Which mode to run (concat recognition or single recognition). Default is False.",
99 | )
100 | videocr_param_group.add_argument(
101 | "-b",
102 | "--batch_size",
103 | type=int,
104 | default=10,
105 | help="The batch of concating image nums in concat recognition mode. Default is 10.",
106 | )
107 |
108 | vsf_param_group = parser.add_argument_group(title="VSFParameters")
109 | vsf_param_group.add_argument(
110 | "-vsf",
111 | "--vsf_exe_path",
112 | type=str,
113 | default=None,
114 | help="The full path of VideoSubFinderWXW.exe.",
115 | )
116 | vsf_param_group.add_argument(
117 | "-c",
118 | "--clear_dirs",
119 | action="store_false",
120 | default=True,
121 | help="Clear Folders (remove all images), performed before any other steps. Default is True",
122 | )
123 | vsf_param_group.add_argument(
124 | "-r",
125 | "--run_search",
126 | action="store_false",
127 | default=True,
128 | help="Run Search (find frames with hardcoded text (hardsub) on video) Default is True",
129 | )
130 | vsf_param_group.add_argument(
131 | "-ccti",
132 | "--create_cleared_text_images",
133 | action="store_true",
134 | default=False,
135 | help="Create Cleared Text Images. Default is True",
136 | )
137 | vsf_param_group.add_argument(
138 | "-ces",
139 | "--create_empty_sub",
140 | type=str,
141 | default=None,
142 | help="Create Empty Sub With Provided Output File Name (*.ass or *.srt)",
143 | )
144 | vsf_param_group.add_argument(
145 | "-cscti",
146 | "--create_sub_from_cleared_txt_images",
147 | type=str,
148 | default=None,
149 | help="Create Sub From Cleared TXT Images With Provided Output File Name (*.ass or *.srt)",
150 | )
151 | vsf_param_group.add_argument(
152 | "-cstxt",
153 | "--create_sub_from_txt_results",
154 | type=str,
155 | default=None,
156 | help="Create Sub From TXT Results With Provided Output File Name (*.ass or *.srt)",
157 | )
158 | vsf_param_group.add_argument(
159 | "-ovocv",
160 | "--open_video_opencv",
161 | action="store_false",
162 | default=True,
163 | help="open video by OpenCV (default). Default is True",
164 | )
165 | vsf_param_group.add_argument(
166 | "-ovffmpeg",
167 | "--open_video_ffmpeg",
168 | action="store_true",
169 | default=False,
170 | help="open video by FFMPEG",
171 | )
172 | vsf_param_group.add_argument(
173 | "-uc", "--use_cuda", action="store_true", default=False, help="use cuda"
174 | )
175 | vsf_param_group.add_argument(
176 | "--start_time",
177 | type=str,
178 | default="0:00:00:000",
179 | help="start time, default = 0:00:00:000 (in format hour:min:sec:milisec)",
180 | )
181 | vsf_param_group.add_argument(
182 | "--end_time",
183 | type=str,
184 | default=None,
185 | help="end time, default = video length",
186 | )
187 | vsf_param_group.add_argument(
188 | "-te",
189 | "--top_video_image_percent_end",
190 | type=float_range(0, 1.0),
191 | default=0.2,
192 | help="top video image percent offset from image bottom, can be in range [0.0,1.0], default = 1.0",
193 | )
194 | vsf_param_group.add_argument(
195 | "-be",
196 | "--bottom_video_image_percent_end",
197 | type=float_range(0, 1.0),
198 | default=0.0,
199 | help="bottom video image percent offset from image bottom, can be in range [0.0,1.0], default = 0.0",
200 | )
201 | vsf_param_group.add_argument(
202 | "-le",
203 | "--left_video_image_percent_end",
204 | type=float_range(0, 1.0),
205 | default=0.0,
206 | help="left video image percent end, can be in range [0.0,1.0], default = 0.0",
207 | )
208 | vsf_param_group.add_argument(
209 | "-re",
210 | "--right_video_image_percent_end",
211 | type=float_range(0, 1.0),
212 | default=1.0,
213 | help="right video image percent end, can be in range [0.0,1.0], default = 1.0",
214 | )
215 | vsf_param_group.add_argument(
216 | "-gs",
217 | "--general_settings",
218 | default=None,
219 | help="general settings (path to general settings *.cfg file, default = settings/general.cfg)",
220 | )
221 | vsf_param_group.add_argument(
222 | "-nthr",
223 | "--num_threads",
224 | type=int,
225 | default=1,
226 | help="number of threads used for Run Search",
227 | )
228 | vsf_param_group.add_argument(
229 | "-nocrthr",
230 | "--num_ocr_threads",
231 | type=int,
232 | default=1,
233 | help="number of threads used for Create Cleared TXT Images",
234 | )
235 | args = parser.parse_args()
236 |
237 | ocr_input_params = RapidVideOCRInput(
238 | is_batch_rec=args.is_batch_rec,
239 | batch_size=args.batch_size,
240 | out_format=args.out_format,
241 | )
242 |
243 | if args.vsf_exe_path and args.video_dir:
244 | vsf_input_params = VideoSubFinderInput(**vars(args))
245 | extractor = RapidVideoSubFinderOCR(vsf_input_params, ocr_input_params)
246 | extractor(args.video_dir, args.save_dir)
247 | elif args.img_dir:
248 | extractor = RapidVideOCR(ocr_input_params)
249 | extractor(args.img_dir, args.save_dir)
250 | else:
251 | pass
252 |
253 |
254 | if __name__ == "__main__":
255 | main()
256 |
--------------------------------------------------------------------------------
/rapid_videocr/ocr_processor.py:
--------------------------------------------------------------------------------
1 | # -*- encoding: utf-8 -*-
2 | # @Author: SWHL
3 | # @Contact: liekkaskono@163.com
4 | from pathlib import Path
5 | from typing import Dict, List, Optional, Tuple
6 |
7 | import cv2
8 | import numpy as np
9 | from rapidocr import RapidOCR
10 | from tqdm import tqdm
11 |
12 | from .utils.logger import logger
13 | from .utils.typings import LOG_LEVEL_MAP
14 | from .utils.utils import (
15 | compute_centroid,
16 | compute_poly_iou,
17 | is_inclusive_each_other,
18 | padding_img,
19 | read_img,
20 | )
21 |
22 |
23 | class OCRProcessor:
24 | def __init__(self, ocr_params: Optional[Dict] = None, batch_size: int = 10):
25 | self.ocr_engine = self._init_ocr_engine(ocr_params)
26 | self.batch_size = batch_size
27 |
28 | def _init_ocr_engine(self, ocr_params: Optional[Dict] = None) -> RapidOCR:
29 | log_level_dict = {"Global.log_level": LOG_LEVEL_MAP[logger.level]}
30 | if ocr_params is None:
31 | return RapidOCR(params=log_level_dict)
32 |
33 | ocr_params.update(log_level_dict)
34 | return RapidOCR(params=ocr_params)
35 |
36 | def __call__(
37 | self, img_list: List[Path], is_batch_rec: bool, is_txt_dir: bool
38 | ) -> Tuple[List[str], List[str], List[str]]:
39 | self.is_txt_dir = is_txt_dir
40 | process_func = self.batch_rec if is_batch_rec else self.single_rec
41 | rec_results = process_func(img_list)
42 | srt_results = self._generate_srt_results(rec_results)
43 | ass_results = self._generate_ass_results(rec_results)
44 | txt_results = self._generate_txt_result(rec_results)
45 | return srt_results, ass_results, txt_results
46 |
47 | def single_rec(self, img_list: List[Path]) -> List[Tuple[int, str, str, str]]:
48 | logger.info("[OCR] Running with single recognition.")
49 |
50 | rec_results = []
51 | for i, img_path in enumerate(tqdm(img_list, desc="OCR")):
52 | time_str = self._get_srt_timestamp(img_path)
53 | ass_time_str = self._get_ass_timestamp(img_path)
54 | img = self._preprocess_image(img_path)
55 |
56 | dt_boxes, rec_res = self.get_ocr_result(img)
57 | txts = (
58 | self.process_same_line(dt_boxes, rec_res)
59 | if dt_boxes is not None
60 | else ""
61 | )
62 | rec_results.append([i, time_str, txts, ass_time_str])
63 | return rec_results
64 |
65 | @staticmethod
66 | def _get_srt_timestamp(file_path: Path) -> str:
67 | """0_00_00_041__0_00_00_415_0070000000019200080001920.jpeg"""
68 |
69 | def format_time(time_parts):
70 | time_parts[0] = f"{time_parts[0]:0>2}"
71 | return ":".join(time_parts[:3]) + f",{time_parts[3]}"
72 |
73 | split_paths = file_path.stem.split("_")
74 | start_time = split_paths[:4]
75 | end_time = split_paths[5:9]
76 | return f"{format_time(start_time)} --> {format_time(end_time)}"
77 |
78 | @staticmethod
79 | def _get_ass_timestamp(file_path: Path) -> str:
80 | s = file_path.stem
81 |
82 | h1 = int(s[0:1])
83 | m1 = int(s[2:4])
84 | sec1 = int(s[5:7])
85 | ms1 = int(s[8:11])
86 |
87 | h2 = int(s[13:14])
88 | m2 = int(s[15:17])
89 | sec2 = int(s[18:20])
90 | ms2 = int(s[21:24])
91 |
92 | # compute absolute times in milliseconds
93 | bt = (h1 * 3600 + m1 * 60 + sec1) * 1000 + ms1
94 | et = (h2 * 3600 + m2 * 60 + sec2) * 1000 + ms2
95 |
96 | def to_ass(ts_ms: int) -> str:
97 | # centiseconds (drop the last digit, no rounding)
98 | cs_total = ts_ms // 10
99 | cs = cs_total % 100
100 | total_s = ts_ms // 1000
101 | s = total_s % 60
102 | total_m = total_s // 60
103 | m = total_m % 60
104 | h = total_m // 60
105 | # H:MM:SS.CC
106 | return f"{h}:{m:02d}:{s:02d}.{cs:02d}"
107 |
108 | return f"{to_ass(bt)},{to_ass(et)}"
109 |
110 | @staticmethod
111 | def _preprocess_image(img_path: Path) -> np.ndarray:
112 | img = read_img(img_path)
113 | img = padding_img(img, (img.shape[0], img.shape[0], 0, 0))
114 | return img
115 |
116 | @staticmethod
117 | def _generate_srt_results(
118 | rec_results: List[Tuple[int, str, str, str]],
119 | ) -> List[str]:
120 | return [f"{i + 1}\n{time_str}\n{txt}\n" for i, time_str, txt, _ in rec_results]
121 |
122 | @staticmethod
123 | def _generate_ass_results(
124 | rec_results: List[Tuple[int, str, str, str]],
125 | ) -> List[str]:
126 | return [
127 | f"Dialogue: 0,{ass_time_str},Default,,0,0,0,,{txt}"
128 | for _, _, txt, ass_time_str in rec_results
129 | ]
130 |
131 | @staticmethod
132 | def _generate_txt_result(rec_results: List[Tuple[int, str, str, str]]) -> List[str]:
133 | return [f"{txt}\n" for _, _, txt, _ in rec_results]
134 |
135 | def batch_rec(self, img_list: List[Path]) -> List[Tuple[int, str, str, str]]:
136 | logger.info("[OCR] Running with concat recognition.")
137 |
138 | img_nums = len(img_list)
139 | rec_results = []
140 | for start_i in tqdm(range(0, img_nums, self.batch_size), desc="Concat Rec"):
141 | end_i = min(img_nums, start_i + self.batch_size)
142 |
143 | concat_img, img_coordinates, img_paths = self._prepare_batch(
144 | img_list[start_i:end_i]
145 | )
146 | dt_boxes, rec_res = self.get_ocr_result(concat_img)
147 | if rec_res is None or dt_boxes is None:
148 | continue
149 |
150 | one_batch_rec_results = self._process_batch_results(
151 | start_i, img_coordinates, dt_boxes, rec_res, img_paths
152 | )
153 | rec_results.extend(one_batch_rec_results)
154 | return rec_results
155 |
156 | def _prepare_batch(
157 | self, img_list: List[Path]
158 | ) -> Tuple[np.ndarray, np.ndarray, List[Path]]:
159 | padding_value = 10
160 | array_img_list, img_coordinates = [], []
161 | for i, img_path in enumerate(img_list):
162 | img = read_img(img_path)
163 | if self.is_txt_dir:
164 | img = cv2.resize(img, None, fx=0.25, fy=0.25)
165 |
166 | pad_img = padding_img(img, (0, padding_value, 0, 0))
167 | array_img_list.append(pad_img)
168 |
169 | h, w = img.shape[:2]
170 | x0, y0 = 0, i * (h + padding_value)
171 | x1, y1 = w, (i + 1) * (h + padding_value)
172 | img_coordinates.append([(x0, y0), (x1, y0), (x1, y1), (x0, y1)])
173 |
174 | return np.vstack(array_img_list), np.array(img_coordinates), img_list
175 |
176 | def _process_batch_results(
177 | self,
178 | start_i: int,
179 | img_coordinates: np.ndarray,
180 | dt_boxes: np.ndarray,
181 | rec_res: Tuple[str],
182 | img_paths: List[Path],
183 | ) -> List[Tuple[int, str, str, str]]:
184 | match_dict = self._match_boxes_to_images(
185 | img_coordinates, dt_boxes, rec_res, img_paths
186 | )
187 |
188 | results = []
189 | for k, v in match_dict.items():
190 | cur_frame_idx = start_i + k
191 | if v:
192 | img_path, boxes, recs = list(zip(*v))
193 | time_str = self._get_srt_timestamp(img_path[0])
194 | ass_time_str = self._get_ass_timestamp(img_path[0])
195 | txts = self.process_same_line(boxes, recs)
196 | else:
197 | time_str = self._get_srt_timestamp(img_paths[k])
198 | ass_time_str = self._get_ass_timestamp(img_paths[k])
199 | txts = ""
200 |
201 | results.append([cur_frame_idx, time_str, txts, ass_time_str])
202 | return results
203 |
204 | def _match_boxes_to_images(
205 | self,
206 | img_coordinates: np.ndarray,
207 | dt_boxes: np.ndarray,
208 | rec_res: List[str],
209 | img_paths: List[Path],
210 | ) -> Dict[int, List[Tuple[Path, np.ndarray, str]]]:
211 | """将检测框匹配到对应图像"""
212 | match_dict = {k: [] for k in range(len(img_coordinates))}
213 | visited_idx = set()
214 |
215 | for i, frame_boxes in enumerate(img_coordinates):
216 | for idx, (dt_box, txt) in enumerate(zip(dt_boxes, rec_res)):
217 | if idx in visited_idx:
218 | continue
219 |
220 | if self._is_box_matched(frame_boxes, dt_box):
221 | match_dict[i].append((img_paths[i], dt_box, txt))
222 | visited_idx.add(idx)
223 |
224 | return match_dict
225 |
226 | def _is_box_matched(self, frame_boxes: np.ndarray, dt_box: np.ndarray) -> bool:
227 | """判断检测框是否匹配到图像"""
228 | box_iou = compute_poly_iou(frame_boxes, dt_box)
229 | return is_inclusive_each_other(frame_boxes, dt_box) or box_iou > 0.1
230 |
231 | def get_ocr_result(
232 | self, img: np.ndarray
233 | ) -> Tuple[Optional[np.ndarray], Optional[Tuple[str]]]:
234 | ocr_result = self.ocr_engine(img)
235 | if ocr_result.boxes is None:
236 | return None, None
237 | return ocr_result.boxes, ocr_result.txts
238 |
239 | def process_same_line(self, dt_boxes: np.ndarray, rec_res: List[str]) -> str:
240 | if len(rec_res) == 1:
241 | return rec_res[0]
242 |
243 | y_centroids = [compute_centroid(box)[1] for box in dt_boxes]
244 | line_groups = self._group_by_lines(y_centroids)
245 | return self._merge_line_text(line_groups, rec_res)
246 |
247 | def _group_by_lines(self, y_centroids: List[float]) -> List[List[int]]:
248 | """将文本框按行分组"""
249 | bool_res = self._is_same_line(y_centroids)
250 | groups = []
251 | current_group = [0]
252 | for i, is_same in enumerate(bool_res, 1):
253 | if is_same:
254 | current_group.append(i)
255 | else:
256 | groups.append(current_group)
257 | current_group = [i]
258 |
259 | groups.append(current_group)
260 | return groups
261 |
262 | @staticmethod
263 | def _is_same_line(points: List) -> List[bool]:
264 | threshold = 5
265 |
266 | align_points = list(zip(points, points[1:]))
267 | bool_res = [False] * len(align_points)
268 | for i, point in enumerate(align_points):
269 | y0, y1 = point
270 | if abs(y0 - y1) <= threshold:
271 | bool_res[i] = True
272 | return bool_res
273 |
274 | def _merge_line_text(self, line_groups: List[List[int]], rec_res: List[str]) -> str:
275 | lines = []
276 | for group in line_groups:
277 | line_text = " ".join(rec_res[i] for i in group)
278 | lines.append(line_text)
279 | return "\n".join(lines)
280 |
--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
1 | Apache License
2 | Version 2.0, January 2004
3 | http://www.apache.org/licenses/
4 |
5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
6 |
7 | 1. Definitions.
8 |
9 | "License" shall mean the terms and conditions for use, reproduction,
10 | and distribution as defined by Sections 1 through 9 of this document.
11 |
12 | "Licensor" shall mean the copyright owner or entity authorized by
13 | the copyright owner that is granting the License.
14 |
15 | "Legal Entity" shall mean the union of the acting entity and all
16 | other entities that control, are controlled by, or are under common
17 | control with that entity. For the purposes of this definition,
18 | "control" means (i) the power, direct or indirect, to cause the
19 | direction or management of such entity, whether by contract or
20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the
21 | outstanding shares, or (iii) beneficial ownership of such entity.
22 |
23 | "You" (or "Your") shall mean an individual or Legal Entity
24 | exercising permissions granted by this License.
25 |
26 | "Source" form shall mean the preferred form for making modifications,
27 | including but not limited to software source code, documentation
28 | source, and configuration files.
29 |
30 | "Object" form shall mean any form resulting from mechanical
31 | transformation or translation of a Source form, including but
32 | not limited to compiled object code, generated documentation,
33 | and conversions to other media types.
34 |
35 | "Work" shall mean the work of authorship, whether in Source or
36 | Object form, made available under the License, as indicated by a
37 | copyright notice that is included in or attached to the work
38 | (an example is provided in the Appendix below).
39 |
40 | "Derivative Works" shall mean any work, whether in Source or Object
41 | form, that is based on (or derived from) the Work and for which the
42 | editorial revisions, annotations, elaborations, or other modifications
43 | represent, as a whole, an original work of authorship. For the purposes
44 | of this License, Derivative Works shall not include works that remain
45 | separable from, or merely link (or bind by name) to the interfaces of,
46 | the Work and Derivative Works thereof.
47 |
48 | "Contribution" shall mean any work of authorship, including
49 | the original version of the Work and any modifications or additions
50 | to that Work or Derivative Works thereof, that is intentionally
51 | submitted to Licensor for inclusion in the Work by the copyright owner
52 | or by an individual or Legal Entity authorized to submit on behalf of
53 | the copyright owner. For the purposes of this definition, "submitted"
54 | means any form of electronic, verbal, or written communication sent
55 | to the Licensor or its representatives, including but not limited to
56 | communication on electronic mailing lists, source code control systems,
57 | and issue tracking systems that are managed by, or on behalf of, the
58 | Licensor for the purpose of discussing and improving the Work, but
59 | excluding communication that is conspicuously marked or otherwise
60 | designated in writing by the copyright owner as "Not a Contribution."
61 |
62 | "Contributor" shall mean Licensor and any individual or Legal Entity
63 | on behalf of whom a Contribution has been received by Licensor and
64 | subsequently incorporated within the Work.
65 |
66 | 2. Grant of Copyright License. Subject to the terms and conditions of
67 | this License, each Contributor hereby grants to You a perpetual,
68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable
69 | copyright license to reproduce, prepare Derivative Works of,
70 | publicly display, publicly perform, sublicense, and distribute the
71 | Work and such Derivative Works in Source or Object form.
72 |
73 | 3. Grant of Patent License. Subject to the terms and conditions of
74 | this License, each Contributor hereby grants to You a perpetual,
75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable
76 | (except as stated in this section) patent license to make, have made,
77 | use, offer to sell, sell, import, and otherwise transfer the Work,
78 | where such license applies only to those patent claims licensable
79 | by such Contributor that are necessarily infringed by their
80 | Contribution(s) alone or by combination of their Contribution(s)
81 | with the Work to which such Contribution(s) was submitted. If You
82 | institute patent litigation against any entity (including a
83 | cross-claim or counterclaim in a lawsuit) alleging that the Work
84 | or a Contribution incorporated within the Work constitutes direct
85 | or contributory patent infringement, then any patent licenses
86 | granted to You under this License for that Work shall terminate
87 | as of the date such litigation is filed.
88 |
89 | 4. Redistribution. You may reproduce and distribute copies of the
90 | Work or Derivative Works thereof in any medium, with or without
91 | modifications, and in Source or Object form, provided that You
92 | meet the following conditions:
93 |
94 | (a) You must give any other recipients of the Work or
95 | Derivative Works a copy of this License; and
96 |
97 | (b) You must cause any modified files to carry prominent notices
98 | stating that You changed the files; and
99 |
100 | (c) You must retain, in the Source form of any Derivative Works
101 | that You distribute, all copyright, patent, trademark, and
102 | attribution notices from the Source form of the Work,
103 | excluding those notices that do not pertain to any part of
104 | the Derivative Works; and
105 |
106 | (d) If the Work includes a "NOTICE" text file as part of its
107 | distribution, then any Derivative Works that You distribute must
108 | include a readable copy of the attribution notices contained
109 | within such NOTICE file, excluding those notices that do not
110 | pertain to any part of the Derivative Works, in at least one
111 | of the following places: within a NOTICE text file distributed
112 | as part of the Derivative Works; within the Source form or
113 | documentation, if provided along with the Derivative Works; or,
114 | within a display generated by the Derivative Works, if and
115 | wherever such third-party notices normally appear. The contents
116 | of the NOTICE file are for informational purposes only and
117 | do not modify the License. You may add Your own attribution
118 | notices within Derivative Works that You distribute, alongside
119 | or as an addendum to the NOTICE text from the Work, provided
120 | that such additional attribution notices cannot be construed
121 | as modifying the License.
122 |
123 | You may add Your own copyright statement to Your modifications and
124 | may provide additional or different license terms and conditions
125 | for use, reproduction, or distribution of Your modifications, or
126 | for any such Derivative Works as a whole, provided Your use,
127 | reproduction, and distribution of the Work otherwise complies with
128 | the conditions stated in this License.
129 |
130 | 5. Submission of Contributions. Unless You explicitly state otherwise,
131 | any Contribution intentionally submitted for inclusion in the Work
132 | by You to the Licensor shall be under the terms and conditions of
133 | this License, without any additional terms or conditions.
134 | Notwithstanding the above, nothing herein shall supersede or modify
135 | the terms of any separate license agreement you may have executed
136 | with Licensor regarding such Contributions.
137 |
138 | 6. Trademarks. This License does not grant permission to use the trade
139 | names, trademarks, service marks, or product names of the Licensor,
140 | except as required for reasonable and customary use in describing the
141 | origin of the Work and reproducing the content of the NOTICE file.
142 |
143 | 7. Disclaimer of Warranty. Unless required by applicable law or
144 | agreed to in writing, Licensor provides the Work (and each
145 | Contributor provides its Contributions) on an "AS IS" BASIS,
146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
147 | implied, including, without limitation, any warranties or conditions
148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
149 | PARTICULAR PURPOSE. You are solely responsible for determining the
150 | appropriateness of using or redistributing the Work and assume any
151 | risks associated with Your exercise of permissions under this License.
152 |
153 | 8. Limitation of Liability. In no event and under no legal theory,
154 | whether in tort (including negligence), contract, or otherwise,
155 | unless required by applicable law (such as deliberate and grossly
156 | negligent acts) or agreed to in writing, shall any Contributor be
157 | liable to You for damages, including any direct, indirect, special,
158 | incidental, or consequential damages of any character arising as a
159 | result of this License or out of the use or inability to use the
160 | Work (including but not limited to damages for loss of goodwill,
161 | work stoppage, computer failure or malfunction, or any and all
162 | other commercial damages or losses), even if such Contributor
163 | has been advised of the possibility of such damages.
164 |
165 | 9. Accepting Warranty or Additional Liability. While redistributing
166 | the Work or Derivative Works thereof, You may choose to offer,
167 | and charge a fee for, acceptance of support, warranty, indemnity,
168 | or other liability obligations and/or rights consistent with this
169 | License. However, in accepting such obligations, You may act only
170 | on Your own behalf and on Your sole responsibility, not on behalf
171 | of any other Contributor, and only if You agree to indemnify,
172 | defend, and hold each Contributor harmless for any liability
173 | incurred by, or claims asserted against, such Contributor by reason
174 | of your accepting any such warranty or additional liability.
175 |
176 | END OF TERMS AND CONDITIONS
177 |
178 | APPENDIX: How to apply the Apache License to your work.
179 |
180 | To apply the Apache License to your work, attach the following
181 | boilerplate notice, with the fields enclosed by brackets "[]"
182 | replaced with your own identifying information. (Don't include
183 | the brackets!) The text should be enclosed in the appropriate
184 | comment syntax for the file format. We also recommend that a
185 | file or class name and description of purpose be included on the
186 | same "printed page" as the copyright notice for easier
187 | identification within third-party archives.
188 |
189 | Copyright [yyyy] [name of copyright owner]
190 |
191 | Licensed under the Apache License, Version 2.0 (the "License");
192 | you may not use this file except in compliance with the License.
193 | You may obtain a copy of the License at
194 |
195 | http://www.apache.org/licenses/LICENSE-2.0
196 |
197 | Unless required by applicable law or agreed to in writing, software
198 | distributed under the License is distributed on an "AS IS" BASIS,
199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
200 | See the License for the specific language governing permissions and
201 | limitations under the License.
202 |
--------------------------------------------------------------------------------
/assets/RapidVideOCRDemo.ipynb:
--------------------------------------------------------------------------------
1 | {
2 | "cells": [
3 | {
4 | "cell_type": "markdown",
5 | "metadata": {
6 | "id": "YBjLERcEsTES"
7 | },
8 | "source": [
9 | "## [RapidVideOCR Demo](https://github.com/SWHL/RapidVideOCR)"
10 | ]
11 | },
12 | {
13 | "cell_type": "markdown",
14 | "metadata": {
15 | "id": "IPBSdGqbjNpc"
16 | },
17 | "source": [
18 | "#### Require:\n",
19 | "- The RGBImages of [Video](https://www.youtube.com/watch?v=Z2Bg_usMYiA) from the VideoSubFinder software.\n",
20 | "- Install the RapidVideOCR"
21 | ]
22 | },
23 | {
24 | "cell_type": "markdown",
25 | "metadata": {
26 | "id": "IT1t_86aq4QU"
27 | },
28 | "source": [
29 | "#### Download the RGBImages.zip and unzip it."
30 | ]
31 | },
32 | {
33 | "cell_type": "code",
34 | "execution_count": 33,
35 | "metadata": {
36 | "colab": {
37 | "base_uri": "https://localhost:8080/"
38 | },
39 | "id": "qWiWiKJWjcH1",
40 | "outputId": "9b5c8098-061a-4f85-b7a8-822e1f26b166"
41 | },
42 | "outputs": [
43 | {
44 | "name": "stdout",
45 | "output_type": "stream",
46 | "text": [
47 | "--2023-04-09 01:47:40-- https://github.com/SWHL/RapidVideOCR/files/11184614/RGBImages.zip\n",
48 | "Resolving github.com (github.com)... 140.82.112.4\n",
49 | "Connecting to github.com (github.com)|140.82.112.4|:443... connected.\n",
50 | "HTTP request sent, awaiting response... 302 Found\n",
51 | "Location: https://objects.githubusercontent.com/github-production-repository-file-5c1aeb/405589029/11184614?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=AKIAIWNJYAX4CSVEH53A%2F20230409%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20230409T014740Z&X-Amz-Expires=300&X-Amz-Signature=fa9c0cb7ec18b1113504c94f60ed8bd6c8250cd040d056396c0dc6caf5184dea&X-Amz-SignedHeaders=host&actor_id=0&key_id=0&repo_id=405589029&response-content-disposition=attachment%3Bfilename%3DRGBImages.zip&response-content-type=application%2Fx-zip-compressed [following]\n",
52 | "--2023-04-09 01:47:40-- https://objects.githubusercontent.com/github-production-repository-file-5c1aeb/405589029/11184614?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=AKIAIWNJYAX4CSVEH53A%2F20230409%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20230409T014740Z&X-Amz-Expires=300&X-Amz-Signature=fa9c0cb7ec18b1113504c94f60ed8bd6c8250cd040d056396c0dc6caf5184dea&X-Amz-SignedHeaders=host&actor_id=0&key_id=0&repo_id=405589029&response-content-disposition=attachment%3Bfilename%3DRGBImages.zip&response-content-type=application%2Fx-zip-compressed\n",
53 | "Resolving objects.githubusercontent.com (objects.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...\n",
54 | "Connecting to objects.githubusercontent.com (objects.githubusercontent.com)|185.199.108.133|:443... connected.\n",
55 | "HTTP request sent, awaiting response... 200 OK\n",
56 | "Length: 8314498 (7.9M) [application/x-zip-compressed]\n",
57 | "Saving to: ‘RGBImages.zip’\n",
58 | "\n",
59 | "RGBImages.zip 100%[===================>] 7.93M 43.3MB/s in 0.2s \n",
60 | "\n",
61 | "2023-04-09 01:47:40 (43.3 MB/s) - ‘RGBImages.zip’ saved [8314498/8314498]\n",
62 | "\n",
63 | "Archive: RGBImages.zip\n",
64 | " creating: RGBImages/\n",
65 | " inflating: RGBImages/0_00_20_640__0_00_23_999_0055800000012800072001280.jpeg \n",
66 | " inflating: RGBImages/0_00_25_120__0_00_25_999_0055800000012800072001280.jpeg \n",
67 | " inflating: RGBImages/0_00_26_000__0_00_26_599_0055800000012800072001280.jpeg \n",
68 | " inflating: RGBImages/0_00_27_760__0_00_28_999_0055800000012800072001280.jpeg \n",
69 | " inflating: RGBImages/0_00_30_280__0_00_30_599_0055800000012800072001280.jpeg \n",
70 | " inflating: RGBImages/0_00_30_600__0_00_32_199_0055800000012800072001280.jpeg \n",
71 | " inflating: RGBImages/0_00_32_800__0_00_33_199_0055800000012800072001280.jpeg \n",
72 | " inflating: RGBImages/0_00_33_200__0_00_34_959_0055800000012800072001280.jpeg \n",
73 | " inflating: RGBImages/0_00_34_960__0_00_35_519_0055800000012800072001280.jpeg \n",
74 | " inflating: RGBImages/0_00_39_040__0_00_39_479_0055800000012800072001280.jpeg \n",
75 | " inflating: RGBImages/0_00_40_040__0_00_41_679_0055800000012800072001280.jpeg \n",
76 | " inflating: RGBImages/0_00_41_680__0_00_42_919_0055800000012800072001280.jpeg \n",
77 | " inflating: RGBImages/0_00_42_920__0_00_43_439_0055800000012800072001280.jpeg \n",
78 | " inflating: RGBImages/0_00_43_440__0_00_43_799_0055800000012800072001280.jpeg \n",
79 | " inflating: RGBImages/0_00_44_920__0_00_45_359_0055800000012800072001280.jpeg \n",
80 | " inflating: RGBImages/0_00_45_360__0_00_47_799_0055800000012800072001280.jpeg \n",
81 | " inflating: RGBImages/0_00_47_800__0_00_48_159_0055800000012800072001280.jpeg \n",
82 | " inflating: RGBImages/0_00_48_160__0_00_48_559_0055800000012800072001280.jpeg \n",
83 | " inflating: RGBImages/0_00_50_520__0_00_54_079_0055800000012800072001280.jpeg \n",
84 | " inflating: RGBImages/0_00_54_080__0_00_55_799_0055800000012800072001280.jpeg \n",
85 | " inflating: RGBImages/0_00_56_000__0_00_56_359_0055800000012800072001280.jpeg \n",
86 | " inflating: RGBImages/0_00_57_680__0_00_57_999_0055800000012800072001280.jpeg \n",
87 | " inflating: RGBImages/0_01_00_120__0_01_00_759_0055800000012800072001280.jpeg \n",
88 | " inflating: RGBImages/0_01_02_160__0_01_02_919_0055800000012800072001280.jpeg \n",
89 | " inflating: RGBImages/0_01_04_320__0_01_06_759_0055800000012800072001280.jpeg \n",
90 | " inflating: RGBImages/0_01_08_120__0_01_08_679_0055800000012800072001280.jpeg \n",
91 | " inflating: RGBImages/0_01_08_680__0_01_13_119_0055800000012800072001280.jpeg \n",
92 | " inflating: RGBImages/0_01_13_120__0_01_13_799_0055800000012800072001280.jpeg \n",
93 | " inflating: RGBImages/0_01_13_800__0_01_16_079_0055800000012800072001280.jpeg \n",
94 | " inflating: RGBImages/0_01_16_080__0_01_17_039_0055800000012800072001280.jpeg \n",
95 | " inflating: RGBImages/0_01_19_320__0_01_20_359_0055800000012800072001280.jpeg \n",
96 | " inflating: RGBImages/0_01_20_360__0_01_21_919_0055800000012800072001280.jpeg \n",
97 | " inflating: RGBImages/0_01_23_120__0_01_23_559_0055800000012800072001280.jpeg \n",
98 | " inflating: RGBImages/0_01_23_560__0_01_24_959_0055800000012800072001280.jpeg \n",
99 | " inflating: RGBImages/0_01_24_960__0_01_25_559_0055800000012800072001280.jpeg \n",
100 | " inflating: RGBImages/0_01_25_560__0_01_26_159_0055800000012800072001280.jpeg \n",
101 | " inflating: RGBImages/0_01_27_560__0_01_27_919_0055800000012800072001280.jpeg \n",
102 | " inflating: RGBImages/0_01_27_920__0_01_30_439_0055800000012800072001280.jpeg \n",
103 | " inflating: RGBImages/0_01_30_440__0_01_31_119_0055800000012800072001280.jpeg \n",
104 | " inflating: RGBImages/0_01_31_120__0_01_31_599_0055800000012800072001280.jpeg \n",
105 | " inflating: RGBImages/0_01_31_600__0_01_32_119_0055800000012800072001280.jpeg \n",
106 | " inflating: RGBImages/0_01_33_040__0_01_34_639_0055800000012800072001280.jpeg \n",
107 | " inflating: RGBImages/0_01_34_640__0_01_38_439_0055800000012800072001280.jpeg \n",
108 | " inflating: RGBImages/0_01_38_440__0_01_38_839_0055800000012800072001280.jpeg \n",
109 | " inflating: RGBImages/0_01_39_960__0_01_40_279_0055800000012800072001280.jpeg \n",
110 | " inflating: RGBImages/0_01_40_280__0_01_40_879_0055800000012800072001280.jpeg \n",
111 | " inflating: RGBImages/0_01_47_920__0_01_48_559_0055800000012800072001280.jpeg \n",
112 | " inflating: RGBImages/0_01_48_560__0_01_50_679_0055800000012800072001280.jpeg \n",
113 | " inflating: RGBImages/0_01_50_920__0_01_51_319_0055800000012800072001280.jpeg \n",
114 | " inflating: RGBImages/0_01_52_520__0_01_53_359_0055800000012800072001280.jpeg \n",
115 | " inflating: RGBImages/0_01_53_360__0_01_53_999_0055800000012800072001280.jpeg \n",
116 | " inflating: RGBImages/0_01_54_000__0_01_56_159_0055800000012800072001280.jpeg \n",
117 | " inflating: RGBImages/0_01_56_160__0_01_56_959_0055800000012800072001280.jpeg \n",
118 | " inflating: RGBImages/0_01_58_040__0_01_58_399_0055800000012800072001280.jpeg \n",
119 | " inflating: RGBImages/0_01_58_400__0_01_59_639_0055800000012800072001280.jpeg \n",
120 | " inflating: RGBImages/0_01_59_640__0_02_00_479_0055800000012800072001280.jpeg \n",
121 | " inflating: RGBImages/0_02_00_480__0_02_01_039_0055800000012800072001280.jpeg \n",
122 | " inflating: RGBImages/0_02_02_240__0_02_02_799_0055800000012800072001280.jpeg \n",
123 | " inflating: RGBImages/0_02_02_800__0_02_04_039_0055800000012800072001280.jpeg \n",
124 | " inflating: RGBImages/0_02_08_000__0_02_09_038_0055800000012800072001280.jpeg \n",
125 | " inflating: RGBImages/0_02_09_039__0_02_10_198_0055800000012800072001280.jpeg \n",
126 | " inflating: RGBImages/0_02_11_720__0_02_13_119_0055800000012800072001280.jpeg \n",
127 | " inflating: RGBImages/0_02_13_280__0_02_13_799_0055800000012800072001280.jpeg \n",
128 | " inflating: RGBImages/0_02_13_800__0_02_14_719_0055800000012800072001280.jpeg \n",
129 | " inflating: RGBImages/0_02_14_720__0_02_15_239_0055800000012800072001280.jpeg \n",
130 | " inflating: RGBImages/0_02_15_240__0_02_15_839_0055800000012800072001280.jpeg \n",
131 | " inflating: RGBImages/0_02_17_640__0_02_21_719_0055800000012800072001280.jpeg \n",
132 | " inflating: RGBImages/0_02_21_720__0_02_22_639_0055800000012800072001280.jpeg \n",
133 | " inflating: RGBImages/0_02_26_640__0_02_27_239_0055800000012800072001280.jpeg \n",
134 | " inflating: RGBImages/0_02_27_240__0_02_27_879_0055800000012800072001280.jpeg \n",
135 | " inflating: RGBImages/0_02_27_920__0_02_28_479_0055800000012800072001280.jpeg \n",
136 | " inflating: RGBImages/0_02_29_360__0_02_30_119_0055800000012800072001280.jpeg \n",
137 | " inflating: RGBImages/0_02_30_240__0_02_30_639_0055800000012800072001280.jpeg \n",
138 | " inflating: RGBImages/0_02_31_200__0_02_31_599_0055800000012800072001280.jpeg \n",
139 | " inflating: RGBImages/0_02_31_600__0_02_32_559_0055800000012800072001280.jpeg \n",
140 | " inflating: RGBImages/0_02_32_560__0_02_33_439_0055800000012800072001280.jpeg \n",
141 | " inflating: RGBImages/0_02_33_440__0_02_34_079_0055800000012800072001280.jpeg \n",
142 | " inflating: RGBImages/0_02_35_520__0_02_37_159_0055800000012800072001280.jpeg \n",
143 | " inflating: RGBImages/0_02_37_160__0_02_41_959_0055800000012800072001280.jpeg \n",
144 | " inflating: RGBImages/0_02_46_440__0_02_47_039_0055800000012800072001280.jpeg \n",
145 | " inflating: RGBImages/0_02_47_040__0_02_48_199_0055800000012800072001280.jpeg \n",
146 | " inflating: RGBImages/0_02_50_520__0_02_50_879_0055800000012800072001280.jpeg \n",
147 | " inflating: RGBImages/0_02_50_880__0_02_53_279_0055800000012800072001280.jpeg \n",
148 | " inflating: RGBImages/0_02_54_840__0_02_56_679_0055800000012800072001280.jpeg \n",
149 | " inflating: RGBImages/0_02_56_680__0_02_57_519_0055800000012800072001280.jpeg \n",
150 | " inflating: RGBImages/0_02_57_520__0_02_57_999_0055800000012800072001280.jpeg \n",
151 | " inflating: RGBImages/0_03_00_360__0_03_00_919_0055800000012800072001280.jpeg \n",
152 | " inflating: RGBImages/0_03_00_920__0_03_01_519_0055800000012800072001280.jpeg \n",
153 | " inflating: RGBImages/0_03_01_560__0_03_04_599_0055800000012800072001280.jpeg \n",
154 | " inflating: RGBImages/0_03_04_600__0_03_05_879_0055800000012800072001280.jpeg \n",
155 | " inflating: RGBImages/0_03_05_880__0_03_06_759_0055800000012800072001280.jpeg \n",
156 | " inflating: RGBImages/0_03_10_160__0_03_10_559_0055800000012800072001280.jpeg \n",
157 | " inflating: RGBImages/0_03_11_680__0_03_11_999_0055800000012800072001280.jpeg \n",
158 | " inflating: RGBImages/0_03_12_040__0_03_12_399_0055800000012800072001280.jpeg \n",
159 | " inflating: RGBImages/0_03_12_400__0_03_12_919_0055800000012800072001280.jpeg \n",
160 | " inflating: RGBImages/0_03_12_920__0_03_13_239_0055800000012800072001280.jpeg \n",
161 | " inflating: RGBImages/0_03_13_240__0_03_13_599_0055800000012800072001280.jpeg \n",
162 | " inflating: RGBImages/0_03_21_000__0_03_21_479_0055800000012800072001280.jpeg \n",
163 | " inflating: RGBImages/0_03_21_600__0_03_21_919_0055800000012800072001280.jpeg \n",
164 | " inflating: RGBImages/0_03_21_920__0_03_22_239_0055800000012800072001280.jpeg \n",
165 | " inflating: RGBImages/0_03_24_480__0_03_24_919_0055800000012800072001280.jpeg \n",
166 | " inflating: RGBImages/0_03_24_920__0_03_25_639_0055800000012800072001280.jpeg \n",
167 | " inflating: RGBImages/0_03_25_640__0_03_27_119_0055800000012800072001280.jpeg \n",
168 | " inflating: RGBImages/0_03_27_120__0_03_27_999_0055800000012800072001280.jpeg \n",
169 | " inflating: RGBImages/0_03_29_520__0_03_30_039_0055800000012800072001280.jpeg \n",
170 | " inflating: RGBImages/0_03_30_120__0_03_30_759_0055800000012800072001280.jpeg \n",
171 | " inflating: RGBImages/0_03_30_760__0_03_31_639_0055800000012800072001280.jpeg \n",
172 | " inflating: RGBImages/0_03_31_640__0_03_31_959_0055800000012800072001280.jpeg \n",
173 | " inflating: RGBImages/0_03_31_960__0_03_32_319_0055800000012800072001280.jpeg \n",
174 | " inflating: RGBImages/0_03_33_680__0_03_33_999_0055800000012800072001280.jpeg \n",
175 | " inflating: RGBImages/0_03_34_000__0_03_34_599_0055800000012800072001280.jpeg \n",
176 | " inflating: RGBImages/0_03_34_600__0_03_35_399_0055800000012800072001280.jpeg \n",
177 | " inflating: RGBImages/0_03_35_520__0_03_37_959_0055800000012800072001280.jpeg \n",
178 | " inflating: RGBImages/0_03_38_400__0_03_38_879_0055800000012800072001280.jpeg \n",
179 | " inflating: RGBImages/0_03_38_880__0_03_39_439_0055800000012800072001280.jpeg \n",
180 | " inflating: RGBImages/0_03_39_440__0_03_39_919_0055800000012800072001280.jpeg \n",
181 | " inflating: RGBImages/0_03_40_160__0_03_40_599_0055800000012800072001280.jpeg \n",
182 | " inflating: RGBImages/0_03_40_600__0_03_40_919_0055800000012800072001280.jpeg \n",
183 | " inflating: RGBImages/0_03_40_920__0_03_41_399_0055800000012800072001280.jpeg \n",
184 | " inflating: RGBImages/0_03_44_240__0_03_44_679_0055800000012800072001280.jpeg \n",
185 | " inflating: RGBImages/0_03_44_680__0_03_44_999_0055800000012800072001280.jpeg \n",
186 | " inflating: RGBImages/0_03_45_000__0_03_49_239_0055800000012800072001280.jpeg \n",
187 | " inflating: RGBImages/0_03_49_240__0_03_50_799_0055800000012800072001280.jpeg \n",
188 | " inflating: RGBImages/0_03_50_840__0_03_51_199_0055800000012800072001280.jpeg \n",
189 | " inflating: RGBImages/0_03_51_200__0_03_51_599_0055800000012800072001280.jpeg \n",
190 | " inflating: RGBImages/0_03_57_240__0_03_57_919_0055800000012800072001280.jpeg \n",
191 | " inflating: RGBImages/0_03_58_440__0_03_59_199_0055800000012800072001280.jpeg \n",
192 | " inflating: RGBImages/0_03_59_200__0_04_05_279_0055800000012800072001280.jpeg \n",
193 | " inflating: RGBImages/0_04_05_280__0_04_06_919_0055800000012800072001280.jpeg \n",
194 | " inflating: RGBImages/0_04_20_840__0_04_21_159_0055800000012800072001280.jpeg \n",
195 | " inflating: RGBImages/0_04_34_720__0_04_35_879_0055800000012800072001280.jpeg \n"
196 | ]
197 | }
198 | ],
199 | "source": [
200 | "!wget https://github.com/SWHL/RapidVideOCR/files/11184614/RGBImages.zip\n",
201 | "!unzip RGBImages.zip\n",
202 | "!rm RGBImages.zip"
203 | ]
204 | },
205 | {
206 | "cell_type": "markdown",
207 | "metadata": {
208 | "id": "1cTofr4Zq_WB"
209 | },
210 | "source": [
211 | "#### Install the RapidVideOCR package."
212 | ]
213 | },
214 | {
215 | "cell_type": "code",
216 | "execution_count": 34,
217 | "metadata": {
218 | "colab": {
219 | "base_uri": "https://localhost:8080/"
220 | },
221 | "id": "5URYsomEqnuh",
222 | "outputId": "3a6093db-bfaa-4069-e92e-2e7bab97f987"
223 | },
224 | "outputs": [
225 | {
226 | "name": "stdout",
227 | "output_type": "stream",
228 | "text": [
229 | "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
230 | "Requirement already satisfied: rapid_videocr in /usr/local/lib/python3.9/dist-packages (2.1.6)\n",
231 | "Requirement already satisfied: rapidocr-onnxruntime>=1.2.2 in /usr/local/lib/python3.9/dist-packages (from rapid_videocr) (1.2.5)\n",
232 | "Requirement already satisfied: tqdm>=4.52.0 in /usr/local/lib/python3.9/dist-packages (from rapid_videocr) (4.65.0)\n",
233 | "Requirement already satisfied: PyYAML in /usr/local/lib/python3.9/dist-packages (from rapidocr-onnxruntime>=1.2.2->rapid_videocr) (6.0)\n",
234 | "Requirement already satisfied: Pillow in /usr/local/lib/python3.9/dist-packages (from rapidocr-onnxruntime>=1.2.2->rapid_videocr) (8.4.0)\n",
235 | "Requirement already satisfied: onnxruntime>=1.7.0 in /usr/local/lib/python3.9/dist-packages (from rapidocr-onnxruntime>=1.2.2->rapid_videocr) (1.14.1)\n",
236 | "Requirement already satisfied: Shapely>=1.7.1 in /usr/local/lib/python3.9/dist-packages (from rapidocr-onnxruntime>=1.2.2->rapid_videocr) (2.0.1)\n",
237 | "Requirement already satisfied: pyclipper>=1.2.1 in /usr/local/lib/python3.9/dist-packages (from rapidocr-onnxruntime>=1.2.2->rapid_videocr) (1.3.0.post4)\n",
238 | "Requirement already satisfied: numpy>=1.19.3 in /usr/local/lib/python3.9/dist-packages (from rapidocr-onnxruntime>=1.2.2->rapid_videocr) (1.22.4)\n",
239 | "Requirement already satisfied: opencv-python>=4.5.1.48 in /usr/local/lib/python3.9/dist-packages (from rapidocr-onnxruntime>=1.2.2->rapid_videocr) (4.7.0.72)\n",
240 | "Requirement already satisfied: six>=1.15.0 in /usr/local/lib/python3.9/dist-packages (from rapidocr-onnxruntime>=1.2.2->rapid_videocr) (1.16.0)\n",
241 | "Requirement already satisfied: sympy in /usr/local/lib/python3.9/dist-packages (from onnxruntime>=1.7.0->rapidocr-onnxruntime>=1.2.2->rapid_videocr) (1.11.1)\n",
242 | "Requirement already satisfied: flatbuffers in /usr/local/lib/python3.9/dist-packages (from onnxruntime>=1.7.0->rapidocr-onnxruntime>=1.2.2->rapid_videocr) (23.3.3)\n",
243 | "Requirement already satisfied: protobuf in /usr/local/lib/python3.9/dist-packages (from onnxruntime>=1.7.0->rapidocr-onnxruntime>=1.2.2->rapid_videocr) (3.20.3)\n",
244 | "Requirement already satisfied: packaging in /usr/local/lib/python3.9/dist-packages (from onnxruntime>=1.7.0->rapidocr-onnxruntime>=1.2.2->rapid_videocr) (23.0)\n",
245 | "Requirement already satisfied: coloredlogs in /usr/local/lib/python3.9/dist-packages (from onnxruntime>=1.7.0->rapidocr-onnxruntime>=1.2.2->rapid_videocr) (15.0.1)\n",
246 | "Requirement already satisfied: humanfriendly>=9.1 in /usr/local/lib/python3.9/dist-packages (from coloredlogs->onnxruntime>=1.7.0->rapidocr-onnxruntime>=1.2.2->rapid_videocr) (10.0)\n",
247 | "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.9/dist-packages (from sympy->onnxruntime>=1.7.0->rapidocr-onnxruntime>=1.2.2->rapid_videocr) (1.3.0)\n"
248 | ]
249 | }
250 | ],
251 | "source": [
252 | "!pip install rapid_videocr"
253 | ]
254 | },
255 | {
256 | "cell_type": "code",
257 | "execution_count": 35,
258 | "metadata": {
259 | "colab": {
260 | "base_uri": "https://localhost:8080/"
261 | },
262 | "id": "13GXToLcrFl8",
263 | "outputId": "d18fb2c0-79ae-4e29-9b27-de7f7e980707"
264 | },
265 | "outputs": [
266 | {
267 | "name": "stdout",
268 | "output_type": "stream",
269 | "text": [
270 | "Running with concat recognition.\n",
271 | "OCR: 100% 14/14 [00:28<00:00, 2.07s/it]\n",
272 | "The file has been saved in the result/result.srt\n",
273 | "The result has been saved to result directory.\n"
274 | ]
275 | }
276 | ],
277 | "source": [
278 | "!rapid_videocr -i /content/RGBImages -s result -o srt"
279 | ]
280 | },
281 | {
282 | "cell_type": "markdown",
283 | "metadata": {
284 | "id": "DR8RbHFisLZK"
285 | },
286 | "source": [
287 | "#### Look the result."
288 | ]
289 | },
290 | {
291 | "cell_type": "code",
292 | "execution_count": 36,
293 | "metadata": {
294 | "colab": {
295 | "base_uri": "https://localhost:8080/"
296 | },
297 | "id": "g2PdZnGJrsdx",
298 | "outputId": "c2107dd0-f099-464c-91a0-247294e69c60"
299 | },
300 | "outputs": [
301 | {
302 | "name": "stdout",
303 | "output_type": "stream",
304 | "text": [
305 | "1\n",
306 | "00:00:20,640 --> 00:00:23,999\n",
307 | "Eyelyinightinmydreams\n",
308 | "\n",
309 | "2\n",
310 | "00:00:25,120 --> 00:00:25,999\n",
311 | "Iseeyou\n",
312 | "\n",
313 | "3\n",
314 | "00:00:26,000 --> 00:00:26,599\n",
315 | "Iseeyou\n",
316 | "\n",
317 | "4\n",
318 | "00:00:27,760 --> 00:00:28,999\n",
319 | "Ifell you\n",
320 | "\n",
321 | "5\n",
322 | "00:00:30,280 --> 00:00:30,599\n",
323 | "That is hiow I know you go on\n",
324 | "\n",
325 | "6\n",
326 | "00:00:30,600 --> 00:00:32,199\n",
327 | "howIknowyougoon\n",
328 | "\n",
329 | "7\n",
330 | "00:00:32,800 --> 00:00:33,199\n",
331 | "That is howIkhowyougo.on\n",
332 | "\n",
333 | "8\n",
334 | "00:00:33,200 --> 00:00:34,959\n",
335 | "That is how I know you go on\n",
336 | "\n",
337 | "9\n",
338 | "00:00:34,960 --> 00:00:35,519\n",
339 | "That is how I know you go on\n",
340 | "\n",
341 | "11\n",
342 | "00:00:40,040 --> 00:00:41,679\n",
343 | "Faracrossthedistance\n",
344 | "\n",
345 | "12\n",
346 | "00:00:41,680 --> 00:00:42,919\n",
347 | "Faracross the distance\n",
348 | "\n",
349 | "13\n",
350 | "00:00:42,920 --> 00:00:43,439\n",
351 | "Faracross thelistance\n",
352 | "\n",
353 | "14\n",
354 | "00:00:43,440 --> 00:00:43,799\n",
355 | "Faracrosshe Mistance\n",
356 | "\n",
357 | "15\n",
358 | "00:00:44,920 --> 00:00:45,359\n",
359 | "and Spaces between us\n",
360 | "\n",
361 | "16\n",
362 | "00:00:45,360 --> 00:00:47,799\n",
363 | "and Spaces between us\n",
364 | "\n",
365 | "17\n",
366 | "00:00:47,800 --> 00:00:48,159\n",
367 | "and Spaces betweenus\n",
368 | "\n",
369 | "18\n",
370 | "00:00:48,160 --> 00:00:48,559\n",
371 | "and Spacesbetween us\n",
372 | "\n",
373 | "19\n",
374 | "00:00:50,520 --> 00:00:54,079\n",
375 | "u havecometo showyou go on\n",
376 | "\n",
377 | "20\n",
378 | "00:00:54,080 --> 00:00:55,799\n",
379 | "You have come to show vou go on\n",
380 | "\n",
381 | "21\n",
382 | "00:00:56,000 --> 00:00:56,359\n",
383 | "You haveoreto show y\n",
384 | "\n",
385 | "23\n",
386 | "00:01:00,120 --> 00:01:00,759\n",
387 | "Near\n",
388 | "\n",
389 | "24\n",
390 | "00:01:02,160 --> 00:01:02,919\n",
391 | "far\n",
392 | "\n",
393 | "25\n",
394 | "00:01:04,320 --> 00:01:06,759\n",
395 | "reveryouare\n",
396 | "\n",
397 | "26\n",
398 | "00:01:08,120 --> 00:01:08,679\n",
399 | "I belieye that the heart does go\n",
400 | "on\n",
401 | "\n",
402 | "27\n",
403 | "00:01:08,680 --> 00:01:13,119\n",
404 | "I believe that the heart does go\n",
405 | "on\n",
406 | "\n",
407 | "28\n",
408 | "00:01:13,120 --> 00:01:13,799\n",
409 | "I believe that the heart does go\n",
410 | "on\n",
411 | "\n",
412 | "29\n",
413 | "00:01:13,800 --> 00:01:16,079\n",
414 | "I believe that the heart does go\n",
415 | "on\n",
416 | "\n",
417 | "30\n",
418 | "00:01:16,080 --> 00:01:17,039\n",
419 | "I believe that the heart does go\n",
420 | "on\n",
421 | "\n",
422 | "31\n",
423 | "00:01:19,320 --> 00:01:20,359\n",
424 | "Once more\n",
425 | "\n",
426 | "32\n",
427 | "00:01:20,360 --> 00:01:21,919\n",
428 | "Once more\n",
429 | "\n",
430 | "33\n",
431 | "00:01:23,120 --> 00:01:23,559\n",
432 | "thedoor\n",
433 | "you\n",
434 | "\n",
435 | "34\n",
436 | "00:01:23,560 --> 00:01:24,959\n",
437 | "you open the door\n",
438 | "\n",
439 | "35\n",
440 | "00:01:24,960 --> 00:01:25,559\n",
441 | "youopen thedoor\n",
442 | "\n",
443 | "36\n",
444 | "00:01:25,560 --> 00:01:26,159\n",
445 | "you open the door\n",
446 | "\n",
447 | "37\n",
448 | "00:01:27,560 --> 00:01:27,919\n",
449 | "And you're here in my heart\n",
450 | "\n",
451 | "38\n",
452 | "00:01:27,920 --> 00:01:30,439\n",
453 | "And you're here in my heart\n",
454 | "\n",
455 | "39\n",
456 | "00:01:30,440 --> 00:01:31,119\n",
457 | "And you're here in iny heart\n",
458 | "\n",
459 | "40\n",
460 | "00:01:31,120 --> 00:01:31,599\n",
461 | "Andyou're here inm heart\n",
462 | "\n",
463 | "41\n",
464 | "00:01:31,600 --> 00:01:32,119\n",
465 | "Andeinmneart\n",
466 | "\n",
467 | "42\n",
468 | "00:01:33,040 --> 00:01:34,639\n",
469 | "my heartwim goonand\n",
470 | "on\n",
471 | "\n",
472 | "43\n",
473 | "00:01:34,640 --> 00:01:38,439\n",
474 | "my heart will go on and\n",
475 | "on\n",
476 | "\n",
477 | "44\n",
478 | "00:01:38,440 --> 00:01:38,839\n",
479 | "my heart will go on and\n",
480 | "on\n",
481 | "\n",
482 | "47\n",
483 | "00:01:47,920 --> 00:01:48,559\n",
484 | "Love can touch us onetime\n",
485 | "\n",
486 | "48\n",
487 | "00:01:48,560 --> 00:01:50,679\n",
488 | "Lovecantouch usonetime\n",
489 | "\n",
490 | "49\n",
491 | "00:01:50,920 --> 00:01:51,319\n",
492 | "Love cantouch usone time\n",
493 | "\n",
494 | "50\n",
495 | "00:01:52,520 --> 00:01:53,359\n",
496 | "And lastforaJifetime\n",
497 | "\n",
498 | "51\n",
499 | "00:01:53,360 --> 00:01:53,999\n",
500 | "And lastfora lifetime\n",
501 | "\n",
502 | "52\n",
503 | "00:01:54,000 --> 00:01:56,159\n",
504 | "And last fora lifetime\n",
505 | "\n",
506 | "53\n",
507 | "00:01:56,160 --> 00:01:56,959\n",
508 | "An st for a lifetime\n",
509 | "\n",
510 | "54\n",
511 | "00:01:58,040 --> 00:01:58,399\n",
512 | "And never let go till\n",
513 | "\n",
514 | "55\n",
515 | "00:01:58,400 --> 00:01:59,639\n",
516 | "And never let go till\n",
517 | "\n",
518 | "56\n",
519 | "00:01:59,640 --> 00:02:00,479\n",
520 | "And never let go till\n",
521 | "\n",
522 | "57\n",
523 | "00:02:00,480 --> 00:02:01,039\n",
524 | "And never let go till\n",
525 | "\n",
526 | "58\n",
527 | "00:02:02,240 --> 00:02:02,799\n",
528 | "we're gone\n",
529 | "\n",
530 | "59\n",
531 | "00:02:02,800 --> 00:02:04,039\n",
532 | "we're gone\n",
533 | "\n",
534 | "62\n",
535 | "00:02:11,720 --> 00:02:13,119\n",
536 | "one true time\n",
537 | "\n",
538 | "65\n",
539 | "00:02:14,720 --> 00:02:15,239\n",
540 | "Tholdto\n",
541 | "\n",
542 | "66\n",
543 | "00:02:15,240 --> 00:02:15,839\n",
544 | "I holdto\n",
545 | "\n",
546 | "67\n",
547 | "00:02:17,640 --> 00:02:21,719\n",
548 | "Imy lifewe'll alwaysgo on\n",
549 | "\n",
550 | "68\n",
551 | "00:02:21,720 --> 00:02:22,639\n",
552 | "I my life we'll always go on\n",
553 | "\n",
554 | "69\n",
555 | "00:02:26,640 --> 00:02:27,239\n",
556 | "Near\n",
557 | "\n",
558 | "70\n",
559 | "00:02:27,240 --> 00:02:27,879\n",
560 | "Near\n",
561 | "\n",
562 | "72\n",
563 | "00:02:29,360 --> 00:02:30,119\n",
564 | "far\n",
565 | "\n",
566 | "74\n",
567 | "00:02:31,200 --> 00:02:31,599\n",
568 | "whereveryou are\n",
569 | "\n",
570 | "75\n",
571 | "00:02:31,600 --> 00:02:32,559\n",
572 | "wherever you-are\n",
573 | "\n",
574 | "76\n",
575 | "00:02:32,560 --> 00:02:33,439\n",
576 | "whereveryou are\n",
577 | "\n",
578 | "77\n",
579 | "00:02:33,440 --> 00:02:34,079\n",
580 | "whereveryou are\n",
581 | "\n",
582 | "78\n",
583 | "00:02:35,520 --> 00:02:37,159\n",
584 | "I believe that the heart does go\n",
585 | "on\n",
586 | "\n",
587 | "79\n",
588 | "00:02:37,160 --> 00:02:41,959\n",
589 | "I believe that the heart does go\n",
590 | "on\n",
591 | "\n",
592 | "80\n",
593 | "00:02:46,440 --> 00:02:47,039\n",
594 | "Once more\n",
595 | "\n",
596 | "81\n",
597 | "00:02:47,040 --> 00:02:48,199\n",
598 | "Oncemore\n",
599 | "\n",
600 | "82\n",
601 | "00:02:50,520 --> 00:02:50,879\n",
602 | "you openthe dooi\n",
603 | "\n",
604 | "83\n",
605 | "00:02:50,880 --> 00:02:53,279\n",
606 | "you open the door\n",
607 | "\n",
608 | "84\n",
609 | "00:02:54,840 --> 00:02:56,679\n",
610 | "And you're here in my heart\n",
611 | "\n",
612 | "85\n",
613 | "00:02:56,680 --> 00:02:57,519\n",
614 | "And you're here in my heart\n",
615 | "\n",
616 | "86\n",
617 | "00:02:57,520 --> 00:02:57,999\n",
618 | "And you're here in my heart\n",
619 | "\n",
620 | "87\n",
621 | "00:03:00,360 --> 00:03:00,919\n",
622 | "heartwill goonand\n",
623 | "\n",
624 | "88\n",
625 | "00:03:00,920 --> 00:03:01,519\n",
626 | "my heart will go on and\n",
627 | "on\n",
628 | "\n",
629 | "89\n",
630 | "00:03:01,560 --> 00:03:04,599\n",
631 | "my heart will go on and\n",
632 | "on\n",
633 | "on\n",
634 | "my heart will go on and\n",
635 | "\n",
636 | "90\n",
637 | "00:03:04,600 --> 00:03:05,879\n",
638 | "on\n",
639 | "\n",
640 | "91\n",
641 | "00:03:05,880 --> 00:03:06,759\n",
642 | "my heart will go on and\n",
643 | "on\n",
644 | "\n",
645 | "102\n",
646 | "00:03:24,920 --> 00:03:25,639\n",
647 | "You're here\n",
648 | "\n",
649 | "103\n",
650 | "00:03:25,640 --> 00:03:27,119\n",
651 | "You're here\n",
652 | "\n",
653 | "104\n",
654 | "00:03:27,120 --> 00:03:27,999\n",
655 | "You're here\n",
656 | "\n",
657 | "105\n",
658 | "00:03:29,520 --> 00:03:30,039\n",
659 | "there's nothing I fear\n",
660 | "\n",
661 | "106\n",
662 | "00:03:30,120 --> 00:03:30,759\n",
663 | "there's nothing Ifear\n",
664 | "\n",
665 | "107\n",
666 | "00:03:30,760 --> 00:03:31,639\n",
667 | "there's nothing Ifear\n",
668 | "\n",
669 | "108\n",
670 | "00:03:31,640 --> 00:03:31,959\n",
671 | "there's nothing I fear\n",
672 | "\n",
673 | "109\n",
674 | "00:03:31,960 --> 00:03:32,319\n",
675 | "there nothigIfear\n",
676 | "\n",
677 | "110\n",
678 | "00:03:33,680 --> 00:03:33,999\n",
679 | "AndIknow\n",
680 | "\n",
681 | "111\n",
682 | "00:03:34,000 --> 00:03:34,599\n",
683 | "AndIknow\n",
684 | "\n",
685 | "112\n",
686 | "00:03:34,600 --> 00:03:35,399\n",
687 | "AndIknow\n",
688 | "\n",
689 | "113\n",
690 | "00:03:35,520 --> 00:03:37,959\n",
691 | "that my heart will you go on\n",
692 | "\n",
693 | "114\n",
694 | "00:03:38,400 --> 00:03:38,879\n",
695 | "that my heart will you go on\n",
696 | "\n",
697 | "115\n",
698 | "00:03:38,880 --> 00:03:39,439\n",
699 | "that my heart will you go on\n",
700 | "\n",
701 | "116\n",
702 | "00:03:39,440 --> 00:03:39,919\n",
703 | "that my heart will you go on\n",
704 | "\n",
705 | "117\n",
706 | "00:03:40,160 --> 00:03:40,599\n",
707 | "that my heart will you go on\n",
708 | "\n",
709 | "118\n",
710 | "00:03:40,600 --> 00:03:40,919\n",
711 | "that my heart will you go on\n",
712 | "\n",
713 | "119\n",
714 | "00:03:40,920 --> 00:03:41,399\n",
715 | "that my heart will you go on\n",
716 | "\n",
717 | "120\n",
718 | "00:03:44,240 --> 00:03:44,679\n",
719 | "Weill stay foreverthsway\n",
720 | "\n",
721 | "121\n",
722 | "00:03:44,680 --> 00:03:44,999\n",
723 | "We'll stayforever this way\n",
724 | "\n",
725 | "122\n",
726 | "00:03:45,000 --> 00:03:49,239\n",
727 | "We'll stay forever this way\n",
728 | "\n",
729 | "123\n",
730 | "00:03:49,240 --> 00:03:50,799\n",
731 | "We'll stay forever this way\n",
732 | "\n",
733 | "124\n",
734 | "00:03:50,840 --> 00:03:51,199\n",
735 | "We'll stay forever this way\n",
736 | "\n",
737 | "125\n",
738 | "00:03:51,200 --> 00:03:51,599\n",
739 | "We'll stay forever this way\n",
740 | "\n",
741 | "126\n",
742 | "00:03:57,240 --> 00:03:57,919\n",
743 | "You are And in my heart\n",
744 | "\n",
745 | "127\n",
746 | "00:03:58,440 --> 00:03:59,199\n",
747 | "my heart will go on and\n",
748 | "on\n",
749 | "\n",
750 | "128\n",
751 | "00:03:59,200 --> 00:04:05,279\n",
752 | "my heart will go on and\n",
753 | "on\n",
754 | "\n",
755 | "129\n",
756 | "00:04:05,280 --> 00:04:06,919\n",
757 | "my heart will go on and\n",
758 | "on\n",
759 | "\n"
760 | ]
761 | }
762 | ],
763 | "source": [
764 | "!cat result/result.srt"
765 | ]
766 | },
767 | {
768 | "cell_type": "code",
769 | "execution_count": null,
770 | "metadata": {
771 | "id": "cNjpqvivs1ZA"
772 | },
773 | "outputs": [],
774 | "source": []
775 | }
776 | ],
777 | "metadata": {
778 | "colab": {
779 | "provenance": []
780 | },
781 | "kernelspec": {
782 | "display_name": "Python 3",
783 | "name": "python3"
784 | }
785 | },
786 | "nbformat": 4,
787 | "nbformat_minor": 0
788 | }
789 |
--------------------------------------------------------------------------------