I am making a python app that process user input from video and detects his hand as he is gonna draw with with his fingers. I have code based on paper "Face Segmentation Using Skin-Color Map in Videophone Applications by Chai and Ngan. def process_hand_image(frame): skin_ycrcb_mint = np.array((0, 133, 77)) skin_ycrcb_maxt = np.array((255, 173, 127)) # Convert the frame to YCrCb color space frame_ycrcb = cv2.cvtColor(frame, cv2.COLOR_BGR2YCR_CB) # Create a binary mask where skin color is detected skin_mask = cv2.inRange(frame_ycrcb, skin_ycrcb_mint, skin_ycrcb_maxt) return skin_mask My goal is to separate hand and skin after that. My idea is to use skeletonization and find the finger tips to track and draw the same as the user do. Can you help me with image processing only solutions no media pipe or CNNs? Continue reading...
