kouwis/Full_Body_LandMarks

Full_Body_LandMarks

Explained

In this repo used Mediapipe solutions in sections:

• Face Mesh
• Hands
• Pose

Together, this will extract all coordinates points for any part of the body. In any class of them, there are two functions:

• Find Points
• Drawing

Find_Points function has some of sub functions those are body parts after that returns a list of class points such as: Face points

1. Divide Find_Points function in class Face into:

   • Left Eye
   • Right Eye
   • Nose
   • Mouth
   • Head

2. Divide Find_Points function in class Hands into:

   • Fingers

3. Divide Find_Points function in class Body into:

   • Chest
   • Arms
   • Legs

Drawing function returns an image with labeled points.

Output

In the face parts, It returns a 2d list the first for all face points and the second for point coordinates otherwise it returns a 3d list the first for all class points, the second for the beginning and end point in the part, and the third for point coordinates

Diagram

Installation

pip install landmark-detection

Usage

#Imports
import landmark as LM
import cv2 as cv

cap = cv.VideoCapture(0)  # 0 as a default webcam
face_detector = LM.Face()  # Make instance from class

while 1:
    _, img = cap.read()

    imgRGB = cv.cvtColor(img, cv.COLOR_BGR2RGB)

    # Any function of them need two parameters (image, imageRGB)
    # color, thickness and circle_radius are default parameters
    face_detector.Drawing(img, imgRGB, color=(0, 110, 100), thickness=3, circle_radius=3)
    face_detector.Find_Points(img, imgRGB)

    # To detect any part of the body you should call (Find_points) function
    # Any function of Them does not need parameters
    print(face_detector.Left_Eye())

    cv.imshow("img", img)
    cv.waitKey(1)