Lightweight singles badminton player posture, court detection tool. The framework has a small amount of code and is easy to modify. You can use this tool to get the shuttle player's pose data in competition easily! It means you'll be able to analyze player play more easily using AI methods.
-
court detect
-
net detect
-
players detect
-
badminton detect(show trajectory)
-
event detect(hitting frame capture)
-
Automatically edits out videos from normal camera viewpoints(show rally)
Installation and instructions are docs/Quick-Start.md.
If you want to know the updating record,please see documents/Update-journal.md.
For more details about the flags and output format, please see docs/run.md.
Automated-Hit-frame-Detection-for-Badminton-Match-Analysis
keypoint_rcnn_training_pytorch
This project is a collaborative effort with Weizhi Tao, supported by RIsports and The Hong Kong Polytechnic University (PolyU).