LDA Based Method for Online Punch Segmentation and Recognition
Shuxu Jing, Howard W. H. Leung, Fazhi He, Taku Komura, Jacky C. P. Chan · 2007
In this paper, we propose an approach for real-time online segmentation and recognition of ten typical types of punches from continuously captured boxing sequences. Complete punch patterns are firstly segmented from continuous boxing sequences. Punches from different subjects, which are of variable arm and body length are normalized and uniformly represented with a vector. Then a LDA based punch type recognizer is applied to the unified punches. The type of a specific punch is determined according to a nearest neighbor rule in the LDA feature space. The proposed approach works efficiently and the recognition accuracy is about 98%. We have implemented a real-time online punch recognition system.