Deep learning has offered robust solutions in image recognition. The aim of this project is to create a sports performance analysis application. It can then reads data from streaming videos and is able to identify players’ positions, poses, shots, etc.
Many algorithms and libraries are introduced and the most effective ones were examined to find the most accurate and fastest, as the main goal of this application is to provide real-time feedback to the players and the coach during gameplay. The model can be trained with any type of sports data. The following image shows a sample image of how the code is successful in identifying the players and rackets although the photo has been shot from an abnormal angle.
For further inquiries about this project please use the contact link.