A STUDY OF THE TAEKWONDO POOMSAE RECOGNITION SYSTEM USED BY MOTION RECOGNITION TECHNIQUES

In the real-world, the most important elements for analysis of human's motions are object's motions. Therefore, many researchers has been researched analysis and retrieved method using motion pictures. But, many technologies focused on analysis of direction of motion or path than recognition of motion. So, we proposed a system which represented and retrieved of object's motion used by motion sequence descriptor and change of motions. An inputted motion data transferred 2-dimensional spatial information used image sequence information. And those data transferred 1-dimensional spatial characterized data used by shape descriptor. And extracted characterized data from motion used shape descriptor as sequential array, and represented association relationships of each objects. Finally, proposed system recognized kind of motion based on Poisson Non-zero probability distribution algorithm. The system which proposed applied not only the Taekwondo Poomsae but also uses from various sports areas. Specially, which system very useful from the sports area which needs a video interpretation and operation of motion recognition, used by search keyword or name.