Jump Shot Detection | YOLO + Mediapipe | Computer Vision Project | Computer at BasketBall
🎯 AI-Powered Basketball Shot Detection | Real-Time Analysis with Pose + Object Detection
Welcome to the future of basketball analytics! 🏀 In this video, we demonstrate an AI-powered system that detects and analyzes jump shots in real time using:
✅ MediaPipe Pose Estimation for detecting player movement
✅ YOLOv8 Object Detection for tracking the basketball, rim, and player
✅ Smart logic to determine shot phases: Stance → Jump → Release → Landing
✅ Real-time tracking of total shots, made shots, and missed shots
✅ Trajectory visualization and automated feedback (✅ SCORED / ❌ MISS)
✅ Exported performance stats to a summary file
✅ Fully annotated output video for review and coaching
🧠 How It Works:
👉 The system detects when a player jumps and releases the ball.
👉 It waits intelligently to see if the ball overlaps with the rim.
👉 If there's contact, it's a made shot; if not after a short delay, it's marked as a miss.
👉The trajectory is drawn, and visual feedback appears instantly on screen.
📊 Perfect for:
🔹 Coaches analyzing shooting form
🔹 Players training solo
🔹 Developers exploring pose + object detection
🔹 Sports AI and computer vision enthusiasts
🔧 Tools & Tech:
🔸 Python, OpenCV
🔸 MediaPipe (Pose Detection)
🔸 YOLOv8 (Object Detection via Ultralytics)
📩 For collaborations or questions: [email protected]