I am looking for someone with expertise and experience in developing computer vision solutions for content analysis of squash matches
As a broad overview; ultimately I am looking for a program that tracks player and ball movements. I am most interested in learning where the ball touches the floor and where the ball is struck and by which player. However, in the short term I am looking for the following:
1) Time Stamped data of where the ball touches the floor on the squash court
2) The ability to manually notate when player a or player b strikes the ball
Once this solution is completed I will then look at adding additional features. Any solution should consider the fact that additional features will be added, so should not be built in a restrictive manner.
The videos that will be analysed will be found on the internet or be matches that I personally record.
Assumptions to take into consideration
1) it will be unlikely that I will be able to control the video angle. The angle may change during the match
2) the dimensions of a squash court are fixed
3) there may be video replays. A solution for this might be a manual notation of when a replay is happening so that it can be excluded from data analysis
4) data should be outputted in a way that is compatible with analysis in Excel, R or Python
To view squash matches; do a youtube search for 'PSA squash tv full matches'
If you are experienced in similar project, then I can assign to you this project immediately.