MediaPipe Pose in Computer Vision

Introduction

MediaPipe Pose helps computers find and track body parts in pictures or videos. It makes it easy to understand human poses without complex setup.

You want to count how many times someone raises their hand in a video.

You want to check if someone is doing an exercise correctly by watching their body movements.

You want to create a game that reacts when a player moves their arms or legs.

You want to analyze dance moves by tracking body positions.

You want to help people with physical therapy by monitoring their posture.

Syntax

Computer Vision

import mediapipe as mp
import cv2

mp_pose = mp.solutions.pose
pose = mp_pose.Pose()

cap = cv2.VideoCapture(0)

while cap.isOpened():
    success, image = cap.read()
    if not success:
        break

    image_rgb = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
    results = pose.process(image_rgb)

    if results.pose_landmarks:
        mp.solutions.drawing_utils.draw_landmarks(
            image, results.pose_landmarks, mp_pose.POSE_CONNECTIONS)

    cv2.imshow('MediaPipe Pose', image)
    if cv2.waitKey(5) & 0xFF == 27:
        break

pose.close()
cap.release()
cv2.destroyAllWindows()

Call pose.process() on an RGB image to get pose landmarks.

Use mp.solutions.drawing_utils.draw_landmarks() to draw the detected pose on the image.

Examples

This example shows how to set up MediaPipe Pose for single images instead of video.

Computer Vision

import mediapipe as mp
mp_pose = mp.solutions.pose
pose = mp_pose.Pose(static_image_mode=True)

# Use static_image_mode=True for single images

This example shows how to access the x coordinate of the left wrist from detected landmarks.

Computer Vision

results = pose.process(image_rgb)
if results.pose_landmarks:
    landmarks = results.pose_landmarks.landmark
    print(f"Left wrist x: {landmarks[mp_pose.PoseLandmark.LEFT_WRIST].x}")

Sample Model

This program captures 10 frames from the webcam and prints if a pose was detected in each frame.

Computer Vision

import mediapipe as mp
import cv2

mp_pose = mp.solutions.pose
pose = mp_pose.Pose()

cap = cv2.VideoCapture(0)

frame_count = 0
while cap.isOpened() and frame_count < 10:
    success, image = cap.read()
    if not success:
        break

    image_rgb = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
    results = pose.process(image_rgb)

    if results.pose_landmarks:
        print(f"Frame {frame_count + 1}: Pose landmarks detected")
    else:
        print(f"Frame {frame_count + 1}: No pose detected")

    frame_count += 1

pose.close()
cap.release()
cv2.destroyAllWindows()

Important Notes

Make sure your camera is connected and accessible before running the code.

MediaPipe Pose works best with clear views of the whole body.

Use static_image_mode=True for processing single images instead of video streams.

Summary

MediaPipe Pose detects body landmarks in images or videos.

It helps track human poses easily without complex setup.

You can use it for fitness, games, or any app needing body movement understanding.