This module perform live video detection using the CNN model with help of cv2 to read the video feed from camera and mideapipe to locate the hand and form ROI(region of interest)
In this model the input was converted into grayscale image and binary filters are applied with appropirate thrusthold for good performance
refer the note book for preprocessing and model architecture and refer demo video for working
video link [https://drive.google.com/file/d/1wbEMmJtIhL8n09e5eyS7liqS9CPaAQ7W/view?usp=sharing]