Soccer Tracking

Full Write-Up

Abstract

I created a computer vision model that identifies players and their teams, referees and the ball and measure key statistics such as possesion, speed and distance traveled. I used YOLO, openCV, and custom functions to deal with the unique challenges of this problem - object identification and classification, multi-track statistics, working with optical flow and perspective transformation

Files

Can be found on my github page link incoming

Credit

Full credit to Code in a Jiff for providing a tutorial for this project