holisticRecognition.py

import cv2
import mediapipe as mp

cap = cv2.VideoCapture(0)

mpHolistic = mp.solutions.holistic
holistic = mpHolistic.Holistic()        #This fastens the recognition. Do not remove!
mpDraw = mp.solutions.drawing_utils
drawing_specs = mpDraw.DrawingSpec(thickness=1,circle_radius=1)

while True:
    success,img = cap.read()
    if not success:
        break
    imgRGB = cv2.cvtColor(img,cv2.COLOR_BGR2RGB)
    results = holistic.process(imgRGB)

    mpDraw.draw_landmarks(img,results.face_landmarks,mpHolistic.FACEMESH_CONTOURS ,drawing_specs,drawing_specs)
    mpDraw.draw_landmarks(img,results.left_hand_landmarks,mpHolistic.HAND_CONNECTIONS,drawing_specs,drawing_specs)
    mpDraw.draw_landmarks(img,results.right_hand_landmarks,mpHolistic.HAND_CONNECTIONS,drawing_specs,drawing_specs)
    mpDraw.draw_landmarks(img,results.pose_landmarks,mpHolistic.POSE_CONNECTIONS,drawing_specs,drawing_specs)
    img = cv2.flip(img, 1)
    cv2.imshow('Image',img)
    if cv2.waitKey(1) & 0xff==ord('q'):
        break