BASIC HAND TRACKING USING PYTHON

    Basically, the hand tracking is based on the 21 landmarks in 3D with multi-hand support, based on high-performance palm detection and hand landmark model.

The 21 hand landmarks will be like, 

And, The output will be like this, 

Packages to install : 

1. pip install opencv-python.
2. pip install mediapipe.

Program : 

# Hand Tracking with python
import cv2
import mediapipe as mp
import time

# Capturing video from the web cam
video = cv2.VideoCapture(0)

# Creating objects for hands
mpHands = mp.solutions.hands
hands = mpHands.Hands()

previousTime = 0
currentTime = 0

while True:
    success, img = video.read()
    # Converting BGR image to RGB image
    RGBimg = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

    # Getting landmark points and storing in multiHandLandmarks variable
    result = hands.process(RGBimg)
    multiHandLandmarks = result.multi_hand_landmarks

    # This 'if' block will run when it's having the handlandmark points
    if multiHandLandmarks:
        for handLandmarks in multiHandLandmarks:
            # Accessing each landmark in the hand
            """
            for index, handLandmark in enumerate(handLandmarks.landmark):
                h, w, c = img.shape  # returns height, width, center
                cx, cy = int(handLandmark.x * w), int(handLandmark.y * h)
                if index:
                    cv2.circle(img, (cx, cy), 10, (255, 0, 255), cv2.FILLED)
            """
            # Drawing and connecting the points
            mp.solutions.drawing_utils.draw_landmarks(img, handLandmarks, mpHands.HAND_CONNECTIONS)

    # calculating frames per second
    currentTime = time.time()
    fps = 1 / (currentTime - previousTime)
    previousTime = currentTime

    # This will show the fps value on screen
    cv2.putText(img, str(int(fps)), (10, 70), cv2.FONT_HERSHEY_COMPLEX, 1, (0, 255, 0), 2)

    cv2.imshow("Image", img)
    cv2.waitKey(1)