Motion Detection Sport Game with OpenCV for Tai Chi

Abstract

The project aims at a webcam motion game - Tai Chi which based on hand’s motion detection. This goal implies how to use hand’s motion detection of real-time response and the use of unconstrained environments. In the game, I will use Open Source Computer Vision Library (OpenCV) to track or recognize hands from real-time webcam images for interacting with the game. In order to capture the hand’s motion, it is based on two main methods: hand segmentation and hand tracking. For the hand segmentation, I will use the color cue due to the characteristic color values of human skin and segment the hand. To tracking the hand motion, Motion History should be used. It is an algorithm that tracks the object moves by calculate the changes of the distribution in the images. By tracking the motions from the players, the game can provide an interactive environment for them to learn the TAI CHI motions.