Author(s)

Shengnan Li, Yan Ma, Hui Huang and Shunbao Li

Corresponding Author

Yan Ma

Abstract

Human motion similarity algorithm is important for human behavior recognition. However, the accuracy and stability of motion similarity algorithm will be affected by different data acquisition objects and object actions. In order to address this issue, a skeleton-adaptive human motion similarity method is proposed. First, the data of human joint points is acquired by Kinect. Second, the joints of the motion to be tested are corrected with the length of template skeleton. Third, the weight of joints are adjusted adaptively. Finally, the similarity between motions is calculated. The experimental results show that the performance of the proposed method is more stable and accurate.

Keywords

Human Behavior Recognition, Kinect, Joint Position Correction, Weight Adaptation, Motion Similarity