Fast and reliable recognition of human motion from motion trajectories using wavelet analysis

Abstract

Recognition of human motion provides hints to understand human activities and gives opportunities to the development of new human-computer interface. Recent studies, however, are limited to extracting motion history image and recognizing gesture or locomotion of human body parts. Although the approach employed, i.e. the transformation of the 3D space-time (x-y-t) analysis to the 2D image analysis, is faster than analyzing 3D motion feature, it is less accurate and less robust in nature. In this paper, a fast trajectory-classification algorithm for interpreting movement of human body parts using wavelet analysis is proposed to increase the accuracy and robustness of human motion recognition. By tracking human body in real time, the motion trajectory (x-y-t) can be extracted. The motion trajectory is then broken down into wavelets that form a set of wavelet features. Classification based on the wavelet features can then be done to interpret the human motion. An online hand drawing digit recognition system was built using the proposed algorithm. Experiments show that the proposed algorithm is able to recognize digits from human movement accurately in real time.