This project gets the advantage of both biomechanics analysis and Kinect motion capturing, and develops a sports improvement solution with coaching evaluation. It focuses on sample movement patterns to do data quantity and quality analysis. And by combining with professional dedicated bio-mechanical principles, it is able to implement real time motion tracking, coaching and evaluation while motion capturing. We calculate some basic but important parameters from captured motion data, such as the rotation and translation of body segments, and then analyze motion flaws that hid behind it. So a deterministic model for specific movement pattern can be constructed as a backend configuration for real time motion coaching. And this deterministic model is the source of motion performance evaluation as well, with which we adopt neural net working to do the scoring job.
Sun, Jing, "Motion Learning with Biomechanics Principles" (2013). Master's Projects. 313.