Sports data analysis

Advances in tracking technology enable us to monitor and analyse match performances of sports persons. This project aims to develop a system that will track, display and analyse movement patterns of sports persons. Available short-range wireless devices reviewed are found to be either unsuitable due to inadequate accuracies or having acceptable accuracies but highly-priced. Hence, an alternative strategy, based on computer vision is adopted. Based on a squash case study, Open Source Computer Vision Library 1.0 is employed to develop the system, Squash Track. Squash Track, tracks squash players motions using avail- able video data and presents players' positional statistic with reference to key areas in the squash court. Based on a test result, where players wore distinguishable attire, Squash Track has an estimated accuracy of 2m for positioning players 80 per cent of the time. Squash Track can be used to monitor players' performance; the positioning accuracy can be further improved and expanded further to analysis squash tactics based on the sports person's position.Bachelor of Engineering (Computer Engineering