dc.description.abstract |
Recognition of players in pictures of sporting events is an approachable but tough task. While there are many good player recognition systems, there are fewer methods that can identify the tracked players. Player identification is challenging in such videos due to blurry facial features and rarely visible jersey numbers (when visible, are deformed due to player movements).In the case of a Cricket game, this task can be accomplished by compartmentalizing the job. By characterizing jersey color with MAP detection, isolating the jerseys and numbers, and using template matching, I tried to make a max-effort algorithm that distinguish however many players as could be allowed. |
en_US |