├── IDrecognition ├── 1.jpg ├── 10.jpg ├── 11.jpg ├── 2.jpg ├── 3.jpg ├── 3.png ├── 4.jpg ├── 5.jpg ├── 6.jpg ├── 7.jpg ├── video.py ├── readme.html └── IDrec.py └── README.md /IDrecognition/1.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/1.jpg -------------------------------------------------------------------------------- /IDrecognition/10.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/10.jpg -------------------------------------------------------------------------------- /IDrecognition/11.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/11.jpg -------------------------------------------------------------------------------- /IDrecognition/2.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/2.jpg -------------------------------------------------------------------------------- /IDrecognition/3.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/3.jpg -------------------------------------------------------------------------------- /IDrecognition/3.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/3.png -------------------------------------------------------------------------------- /IDrecognition/4.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/4.jpg -------------------------------------------------------------------------------- /IDrecognition/5.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/5.jpg -------------------------------------------------------------------------------- /IDrecognition/6.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/6.jpg -------------------------------------------------------------------------------- /IDrecognition/7.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/lvxiaojie111/OCR/HEAD/IDrecognition/7.jpg -------------------------------------------------------------------------------- /IDrecognition/video.py: -------------------------------------------------------------------------------- 1 | ############################################################################### 2 | #createor: junjielv199108 3 | #reate time: 2020/10/01/15/59 4 | ############################################################################### 5 | import cv2 6 | import numpy 7 | cap = cv2.VideoCapture(0)# 调整参数实现读取视频或调用摄像头 8 | while 1: 9 | ret, frame = cap.read() 10 | try: 11 | if ret: 12 | cv2.imshow("cap", frame) 13 | if cv2.waitKey(100) & 0xff == ord('q'): 14 | break; 15 | except: 16 | print(1) 17 | cap.release() 18 | cv2.destroyAllWindows() -------------------------------------------------------------------------------- /IDrecognition/readme.html: -------------------------------------------------------------------------------- 1 | crater:junjielv 2020/10/04 2 | 2020/10/4 身份证识别项目 ----v1.0 3 | 1、所需库：pytesseract.py 4 | 安装：pip install pytesseract 5 | 2、运行video.py代码，调试在线摄像头 6 | 3、运行IDrec.py,进行身份证识别 7 | 功能1：图片处理后识别 8 | 功能2：图片直接识别 9 | 功能3：视频在线识别 10 | 11 | 2020/10/5 12 | 加： 13 | 1、中文文字识别（下载chi_sim.traineddata中文文字库：https://tesseract-ocr.github.io/tessdoc/Data-Files） 14 | github官方很慢，可参考：https://pan.baidu.com/s/1uuSTBNo3byJib4f8eRSIFw ,密：8v8u 15 | 可参考：https://blog.csdn.net/qq_38161040/article/details/90668765 16 | 17 | 18 | 常见错误： 19 | 1）tesseract is not installed or it's not in your path 20 | ①需要在 window本地环境下安装：Tesseract-OCR 下载地址：https://github.com/tesseract-ocr/tesseract/wiki 21 | ②解压安装，并设置环境变量，可参考：https://blog.csdn.net/weixin_40569991/article/details/82082173 22 | ③打开anaconda\envs\yourvirenv\site-packages\pytesseract\pytesseract.py文件;将tesseract_cmd = r'd\Tesseract-OCR\tesseract.exe' 23 | ④重新在python中运行Idrec.py即可成功. 24 | 25 | 优点： 26 | 1、对身份证件的识别要好一些 27 | 2、对标准的大字体识别好些 28 | 29 | 缺点： 30 | 1、对广告拍字体的识别 效果特别差 31 | 2、对车牌的识别也特别差 32 | 33 | 34 | 35 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # OCR 2 | 身份证文字识别项目 3 | 4 | crater:junjielv 2020/10/04 5 | 2020/10/4 身份证识别项目 ----v1.0 6 | 7 | 1、所需库：pytesseract.py 8 | 安装：pip install pytesseract 9 | 10 | 2、运行video.py代码，调试在线摄像头 11 | 12 | 3、运行IDrec.py,进行身份证识别 13 | 14 | 功能1：图片处理后识别 15 | 16 | 功能2：图片直接识别 17 | 18 | 功能3：视频在线识别 19 | 20 | 2020/10/5 21 | 加： 22 | 23 | 1、中文文字识别（下载chi_sim.traineddata中文文字库：https://tesseract-ocr.github.io/tessdoc/Data-Files） 24 | github官方很慢，可参考：https://pan.baidu.com/s/1uuSTBNo3byJib4f8eRSIFw ,密：8v8u 25 | 可参考：https://blog.csdn.net/qq_38161040/article/details/90668765 26 | 27 | 28 | 常见错误： 29 | 30 | 1）tesseract is not installed or it's not in your path 31 | 32 | ①需要在 window本地环境下安装：Tesseract-OCR 下载地址：https://github.com/tesseract-ocr/tesseract/wiki 33 | ②解压安装，并设置环境变量，可参考：https://blog.csdn.net/weixin_40569991/article/details/82082173 34 | ③打开anaconda\envs\yourvirenv\site-packages\pytesseract\pytesseract.py文件;将tesseract_cmd = r'd\Tesseract-OCR\tesseract.exe' 35 | ④重新在python中运行Idrec.py即可成功. 36 | 37 | 优点： 38 | 39 | 1、对身份证件的识别要好一些 40 | 2、对标准的大字体识别好些 41 | 42 | 缺点： 43 | 1、对广告拍字体的识别 效果特别差 44 | 2、对车牌的识别也特别差 45 | -------------------------------------------------------------------------------- /IDrecognition/IDrec.py: -------------------------------------------------------------------------------- 1 | # !/usr/bin/python 2 | #-*-coding:utf-8-*- 3 | import sys 4 | # import importlib 5 | # importlib.reload(sys) 6 | # sys.setdefaultencoding('utf-8') 7 | #createor: junjielv199108 8 | #create time: 2020/10/01/15/59 9 | import time 10 | time1 = time.time() 11 | from PIL import Image 12 | import pytesseract#ocr字符识别库 13 | import cv2 14 | ###########二值化算法 15 | def binarizing(img,threshold): 16 | pixdata = img.load() 17 | w, h = img.size 18 | for y in range(h): 19 | for x in range(w): 20 | if pixdata[x, y] < threshold: 21 | pixdata[x, y] = 0 22 | else: 23 | pixdata[x, y] = 255 24 | return img 25 | 26 | ###########去除干扰线算法 27 | def depoint(img): #input: gray image 28 | pixdata = img.load() 29 | w,h = img.size 30 | for y in range(1,h-1): 31 | for x in range(1,w-1): 32 | count = 0 33 | if pixdata[x,y-1] > 245: 34 | count = count + 1 35 | if pixdata[x,y+1] > 245: 36 | count = count + 1 37 | if pixdata[x-1,y] > 245: 38 | count = count + 1 39 | if pixdata[x+1,y] > 245: 40 | count = count + 1 41 | if count > 2: 42 | pixdata[x,y] = 255 43 | return img 44 | ########身份证号码识别 45 | def identity_OCR(pic_path): 46 | #####身份证号码截图 47 | img1=Image.open(pic_path) 48 | w,h=img1.size 49 | ##将身份证放大3倍 50 | out=img1.resize((w*3,h*3),Image.ANTIALIAS) 51 | Image._show(out) 52 | cv2.waitKey(0) 53 | region = (125*3,200*3,370*3,250*3) 54 | #裁切身份证号码图片 55 | cropImg = out.crop(region) 56 | # print(cropImg) 57 | # Image._show(cropImg) 58 | # cv2.waitKey(0) 59 | # 转化为灰度图 60 | img= cropImg.convert('L') 61 | Image._show(img) 62 | cv2.waitKey(0) 63 | # 把图片变成二值图像。 64 | img1=binarizing(img,100) 65 | img2=depoint(img) 66 | code = pytesseract.image_to_string(img2) 67 | print("识别该身份证号码是:"+str(code)) 68 | ''' 69 | ########身份证号码识别################### 70 | ##########英文图片识别#################### 71 | ''' 72 | def identity_OCR_Nopro(pic_path): 73 | image = Image.open(pic_path) 74 | content = pytesseract.image_to_string(image) # 解析图片 75 | print(content) 76 | ''' 77 | #####################中文图片识别############################### 78 | 要在pytesseract 库的 image_to_string() 方法里加个参数lang='chi_sim'， 79 | 这个就是引用对应的中文语言包，中文语言包的全名是 chi_sim.traineddata 80 | ''' 81 | def identity_OCR_Chine(pic_path): 82 | from PIL import Image 83 | import pytesseract 84 | image = Image.open(pic_path) 85 | content = pytesseract.image_to_string(image, lang='chi_sim') # 解析图片 86 | print(content) 87 | 88 | ''' 89 | ##############功能：video recognition############# 90 | ''' 91 | def identity_OCR_Video(pic_path): 92 | vid = cv2.VideoCapture(pic_path) 93 | while True: 94 | try: 95 | return_value, frame = vid.read() 96 | if return_value: 97 | cv2.imshow("result", frame) 98 | frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB) 99 | 100 | # image = Image.fromarray(frame) 101 | # result = cv2.cvtColor(image, cv2.COLOR_RGB2BGR) 102 | code = pytesseract.image_to_string(frame,lang='chi_sim') 103 | print("识别该身份证号码是:" + str(code)) 104 | cv2.waitKey(100) 105 | else: 106 | raise ValueError("No image!") 107 | except: 108 | print() 109 | if __name__ == '__main__': 110 | pic_path="./11.jpg" 111 | # identity_OCR(pic_path) 112 | # identity_OCR_Video(0) 113 | identity_OCR_Chine(pic_path) 114 | # identity_OCR_Nopro(pic_path) 115 | 116 | time2 = time.time() 117 | print(u'总共耗时：' + str(time2 - time1) + 's') 118 | 119 | 120 | --------------------------------------------------------------------------------