Hand Tracking with opencv and mediapipe on python

Simple hand tracking using opencv and mediapipe.

I start from the code of opencvhandracking library that actually do not work on my python 3.6 and 3.8. I copy their classes and I use it into a very simple script.

You need to have python and pip and install the following libraries

pip install mediapipe opencv-python

The code can be downloaded here

Now connect a webcam and exec it with:

python3 main.py

As in the figure you will see the output of your webcam and the hand tracking draw. Press ‘q’ to quit.

This code is very simple and use only 51 lines of codes, but when the library will work well with pip it can be replaced with 10 as follow

import cv2
from opencvtracking import HandGesture
cap = cv2.VideoCapture(0)
handtrckingdetector = HandGesture()
while True:
success, img = cap.read()
img = handtrckingdetector.drawHand(img)
# landmarkList = handtrckingdetector.handPosition(img)

cv2.imshow("Image", img)
if cv2.waitKey(1) & 0xFF == ord('q'):

This can be very useful for to much applications, for example to create interactive museum objects.

Linux user and Open Source fun. Deep learning PhD Student, Full stack web developer, Mobile developer