Motion & Body Movement Detector
Python Computer Vision
01 — Project Overview
A Python computer vision system using OpenCV and MediaPipe for motion detection and full-body skeletal tracking.
Solo Developer.
Stack
02 — The Problem
Goal
Explore real-time computer vision beyond simple image classification.
User Issue
Casual exploration of MediaPipe's pose estimation for body movement recognition.
03 — The Process
MediaPipe Pose
Used Google's MediaPipe Pose solution for 33-landmark full-body skeleton tracking without training a custom model.
Key Challenge
Combining frame-diff motion detection with pose tracking without the two interfering with each other.
results = pose.process(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))
if results.pose_landmarks:
mp_drawing.draw_landmarks(frame, results.pose_landmarks, mp_pose.POSE_CONNECTIONS)04 — The Solution
A real-time Python system that overlays skeleton landmarks on a live camera feed and flags motion events.
05 — Results & Learnings
Impact
Successfully demonstrated real-time 33-point body landmark tracking at ~30fps on a standard laptop webcam.
Learnings
Learned the practical limits of CPU-based MediaPipe vs GPU-accelerated approaches.
What I'd do differently
Would add gesture recognition to map body poses to custom keyboard shortcuts.