bodycr 1.0.0

Body Capture and Recognition

Body CR - Body Capture and Recognition Library

The Body CR make easy the body recognition using OpenCv and Mediapipe, providing an easy interface to detect complete body and hands

Usage

To use the Pose detector, just import the library with import BodyCR as cr and create a new instance of the Capture class using capture = cr.Recognize(). Now, to every frame capture, you need read all landmarks using capture.Read(img), all landmarks was generated is in capture.pose (to pose detection), capture.face (to pfaceose detection) and capture.hands (to hand detection, in all detection, the index 0 in the right hand and 1 is left hand).

Full example

# Importing OpenCV and BodyCR
import cv2
import BodyCR as cr

cap = cv2.VideoCapture(0) # Creating cv2 video capture

# Setting up the BodyCR Workspace

## Creating Base Capture
capture = cr.Recognize.Capture(
    pose=cr.Prefabs.POSE.normal.Mount(),
)
## Creating Drawer
draw = cr.Drawer()
## Creating the FPS Manager
fps = cr.FPS()

# Main Loop
while True:
    _, img = cap.read() # Reading the camera
    img = cv2.flip(img, 1) # Fliping the image

    draw.UpdateImage(img) # Update inset drawer image
    capture.Read(img, cr.Prefabs.DETECT_POSE) # Read the image with the BodyCR Capture

    draw.DrawComponent(capture.pose, cr.Pose.PoseLandmarks.POSE_CONNECTIONS) # Drawing the connections with Drawer

    fps.Update(img) # Update FPS Image
    cv2.imshow("README Test", img) # Show the result image

    if cv2.waitKey(1) & 0xFF == ord("q"):
        break

cv2.destroyAllWindows()