Tennis · Tool
BTC Computer Vision Tagger
2025-26Updated Aug 21, 2024
Project Overview
A computer-vision tagging system for UCLA Tennis match video that automatically tracks court geometry, player movement, and rally context to support downstream match statistics and reports.
Problem & Objective
UCLA Tennis needed a more efficient way to collect structured data from match videos instead of relying on fully manual tagging workflows.
Analytical Approach
The project adapts an open-source tennis computer-vision pipeline with UCLA-specific homography and camera-system adjustments, then processes model output into data formats usable for match statistics and reporting.
Key Result & Impact
Produced a working automated tagger demo that overlays court lines, player detections, and distance measurements on tennis video for review and analysis.
Tech Stack & Tools
PythonComputer VisionHomographyVideo Processing
Demo video