Computer Virtual Technology for Smart Martial Arts Training based on Body Action Recognition Algorithms

Computer virtual technology for the smart martial arts training based on body action recognition algorithms is the focus of this study. The proposed algorithm clustered the training samples, obtained M cluster centers, and calculated the variance of each dimension of the corresponding clusters, and used them as the initialization parameters of the fuzzy classifier. All the operations in the virtual machine will not affect the host environment, especially the damage to the hardware equipment is completely avoided. Furthermore, the designed model is then applied to the smart martial arts training. The whole pipeline is designed and tested to validate the performance.