Computer vision experiment · Computer vision experiment
Padel Vision
A monocular match-analysis pipeline that tracks four players and the ball, projects play onto a court minimap, and renders annotated video.
YOLOv11 PyTorch OpenCV Pose Estimation Homography
The Problem
Broadcast footage provides one moving camera view, while useful match analysis needs stable player, ball, and court coordinates.
The Solution
Adjie built a video pipeline combining fine-tuned ball detection, player pose tracking, court homography, and match overlays.
Key Capabilities & Workflow
- Detects and tracks the ball at high resolution.
- Tracks four players with pose estimation.
- Projects detections through court homography.
- Renders trajectories, identities, speed estimates, and match annotations.
Verified Evidence
Input
Single broadcast camera
Output
Annotated video · court minimap
Why It Matters
The experiment turns ordinary match footage into inspectable spatial and motion data without a multi-camera setup.
Confidentiality & Limitations: The portfolio presents an experiment and public media, not a production analytics service.