├── HandTrackingModule.py ├── Readme.md ├── img ├── click.png └── result.gif ├── libraries.bat ├── main.py └── requirements.txt /HandTrackingModule.py: -------------------------------------------------------------------------------- 1 | import cv2 2 | import mediapipe as mp 3 | import time 4 | import math 5 | 6 | class handDetector(): 7 | def __init__(self, mode=False, maxHands=2, modelComplexity=1, detectionCon=0.5, trackCon=0.5): 8 | self.mode = mode 9 | self.maxHands = maxHands 10 | self.modelComplexity = modelComplexity 11 | self.detectionCon = detectionCon 12 | self.trackCon = trackCon 13 | 14 | self.mpHands = mp.solutions.hands 15 | self.hands = self.mpHands.Hands(self.mode, self.maxHands, self.modelComplexity, self.detectionCon, self.trackCon) 16 | self.mpDraw = mp.solutions.drawing_utils 17 | self.tipIds = [4, 8, 12, 16, 20] 18 | 19 | def findHands(self, img, draw=True): 20 | imgRGB = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) 21 | self.results = self.hands.process(imgRGB) 22 | #print(results.multi_hand_landmarks) 23 | 24 | if self.results.multi_hand_landmarks: 25 | for handLms in self.results.multi_hand_landmarks: 26 | if draw: 27 | self.mpDraw.draw_landmarks(img, handLms, self.mpHands.HAND_CONNECTIONS) 28 | return img 29 | 30 | def findPosition(self, img, handNo=0, draw=True): 31 | xList = [] 32 | yList = [] 33 | bbox = [] 34 | self.lmList = [] 35 | if self.results.multi_hand_landmarks: 36 | myHand = self.results.multi_hand_landmarks[handNo] 37 | for id, lm in enumerate(myHand.landmark): 38 | #print(id, lm) 39 | h, w, c = img.shape 40 | cx, cy = int(lm.x*w), int(lm.y*h) 41 | xList.append(cx) 42 | yList.append(cy) 43 | #print(id, cx, cy) 44 | self.lmList.append([id, cx, cy]) 45 | if draw: 46 | cv2.circle(img, (cx, cy), 5, (255,0,255), cv2.FILLED) 47 | xmin, xmax = min(xList), max(xList) 48 | ymin, ymax = min(yList), max(yList) 49 | bbox = xmin, ymin, xmax, ymax 50 | 51 | if draw: 52 | cv2.rectangle(img, (bbox[0]-20, bbox[1]-20), (bbox[2]+20, bbox[3]+20), (0, 255, 0), 2) 53 | return self.lmList, bbox 54 | 55 | def findDistance(self, p1, p2, img, draw=True): 56 | x1, y1 = self.lmList[p1][1], self.lmList[p1][2] 57 | x2, y2 = self.lmList[p2][1], self.lmList[p2][2] 58 | cx, cy = (x1+x2)//2, (y1+y2)//2 59 | 60 | if draw: 61 | cv2.circle(img, (x1,y1), 15, (255,0,255), cv2.FILLED) 62 | cv2.circle(img, (x2,y2), 15, (255,0,255), cv2.FILLED) 63 | cv2.line(img, (x1,y1), (x2,y2), (255,0,255), 3) 64 | cv2.circle(img, (cx,cy), 15, (255,0,255), cv2.FILLED) 65 | 66 | length = math.hypot(x2-x1, y2-y1) 67 | return length, img, [x1, y1, x2, y2, cx, cy] 68 | 69 | def fingersUp(self): 70 | fingers = [] 71 | 72 | # Thumb 73 | if self.lmList[self.tipIds[0]][1] < self.lmList[self.tipIds[0]-1][1]: 74 | fingers.append(1) 75 | else: 76 | fingers.append(0) 77 | 78 | # 4 Fingers 79 | for id in range(1,5): 80 | if self.lmList[self.tipIds[id]][2] < self.lmList[self.tipIds[id]-2][2]: 81 | fingers.append(1) 82 | else: 83 | fingers.append(0) 84 | return fingers 85 | 86 | def main(): 87 | pTime = 0 88 | cTime = 0 89 | cap = cv2.VideoCapture(0) 90 | detector = handDetector() 91 | while True: 92 | success, img = cap.read() 93 | img = detector.findHands(img) 94 | lmList = detector.findPosition(img) 95 | if len(lmList) != 0: 96 | print(lmList[1]) 97 | 98 | cTime = time.time() 99 | fps = 1. / (cTime - pTime) 100 | pTime = cTime 101 | 102 | cv2.putText(img, str(int(fps)), (10,70), cv2.FONT_HERSHEY_PLAIN, 3, (255,0,255), 3) 103 | 104 | cv2.imshow("Image", img) 105 | cv2.waitKey(1) 106 | 107 | 108 | if __name__ == "__main__": 109 | main() -------------------------------------------------------------------------------- /Readme.md: -------------------------------------------------------------------------------- 1 | # Drag and Drop 2 | In this project I am going to learn how to create a virtual drag and drop system using opencv and python 3 | 4 | ## How to install 5 | 1. Clone this repository on your computer 6 | `https://github.com/paveldat/drag_and_drop.git` 7 | 2. Install all the requirements 8 | `run libraries.bat` or 9 | `pip install -r requirements.txt` 10 | 3. Run the program 11 | `python main.py` 12 | 13 | ## Help 14 | You might face issue with webcam not showing and you get errors. 15 | To solve it just change the value in this line (for example to `1`). 16 | `cap = cv2.VideoCapture(0)` 17 | Increment this number until you see your webcam. 18 | 19 | ## Hand Landmarks 20 | 21 | 22 | ## Click 23 | In order to simulate a click, you need to connect the index and middle fingers on your hand. An example of a valid click is shown in the image below. 24 | 25 | 26 | 27 | ## Result 28 | ![Alt Text](https://github.com/paveldat/drag_and_drop/blob/main/img/result.gif) -------------------------------------------------------------------------------- /img/click.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/paveldat/drag_and_drop/41d7cc1dce6950ddaeb5a0bef510730ddc632215/img/click.png -------------------------------------------------------------------------------- /img/result.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/paveldat/drag_and_drop/41d7cc1dce6950ddaeb5a0bef510730ddc632215/img/result.gif -------------------------------------------------------------------------------- /libraries.bat: -------------------------------------------------------------------------------- 1 | pip install opencv-python 2 | pip install numpy 3 | pip install cvzone -------------------------------------------------------------------------------- /main.py: -------------------------------------------------------------------------------- 1 | import cv2 2 | import HandTrackingModule as htm 3 | import cvzone 4 | import numpy as np 5 | 6 | cap = cv2.VideoCapture(0) 7 | cap.set(3, 1280) 8 | cap.set(4, 720) 9 | detector = htm.handDetector(detectionCon=0.8) 10 | colorR = (255, 0, 255) 11 | 12 | cx, cy, w, h = 100, 100, 200, 200 13 | 14 | class DragAndDrogRectangle(): 15 | def __init__(self, posCenter, size=[200, 200]): 16 | self.posCenter = posCenter 17 | self.size = size 18 | 19 | def update(self, cursor): 20 | cx, cy = self.posCenter 21 | w, h = self.size 22 | 23 | # If the index finger tip is in rectangle region 24 | if cx - w//2 < cursor[0] < cx + w//2 and cy - h//2 < cursor[1] < cy + h//2: 25 | self.posCenter = cursor 26 | 27 | rectList = [] 28 | for x in range(5): 29 | rectList.append(DragAndDrogRectangle([x*250+150, 150])) 30 | 31 | while True: 32 | success, img = cap.read() 33 | img = cv2.flip(img, 1) 34 | img = detector.findHands(img) 35 | lmList, _ = detector.findPosition(img) 36 | 37 | if len(lmList) != 0: 38 | length, _, _ = detector.findDistance(8, 12, img, draw=False) 39 | if length < 40: 40 | cursor = lmList[8][1:] # index finger tip landmark 41 | # call the update 42 | for rect in rectList: 43 | rect.update(cursor) 44 | 45 | # Draw 46 | imgNew = np.zeros_like(img, np.uint8) 47 | for rect in rectList: 48 | cx, cy = rect.posCenter 49 | w, h = rect.size 50 | cv2.rectangle(imgNew, (cx - w//2, cy - h//2), (cx + w//2, cy + h//2), colorR, cv2.FILLED) 51 | cvzone.cornerRect(imgNew, (cx - w//2, cy - h//2, w, h), 20, rt=0) 52 | 53 | out = img.copy() 54 | alpha = 0.1 55 | mask = imgNew.astype(bool) 56 | out[mask] = cv2.addWeighted(img, alpha, imgNew, 1-alpha, 0)[mask] 57 | 58 | cv2.imshow("Image", out) 59 | cv2.waitKey(1) -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | opencv-python 2 | numpy 3 | cvzone --------------------------------------------------------------------------------