"Football Player Tracking"

Supervised by Dr Y Gotoh

Abstract

Football is a ball game played between two teams, each attempting to win by scoring more goals than their opponent. Nowadays people enjoy watching football matches not only thought it as a popular sport game but more regard it as a kind of culture and spirit.

The aim the project is to utilize the techniques from Computer Visions to detect and retrieve the track of the football player from computer videos. In general, the whole procession can be divided into five main parts: Recourse Analysis, Video Processing, Image Processing, Data Collection and Modeling Rebuild.

The whole system was regarded as a multi-media information retrieve system rather than a pure image processing system. Thus, it considered more issues during system design including a simple network architecture that can provide a concurrent processing. However, the main system is still concerned on image analysis and processing by image dividing techniques that developed in this project.