Journal of System Simulation ›› 2022, Vol. 34 ›› Issue (3): 633-639.doi: 10.16182/j.issn1004731x.joss.20-0767

• VV&A Technology • Previous Articles     Next Articles

Research on Real-Time Motion Matching of Shadow Play Based on Kinect

Chuanqian Tang(), Zhiqiang Liu, Yijun Su, Xiaojing Liu   

  1. Department of Computer Technology and Application, Qinghai University, Xining 810016, China
  • Received:2020-10-10 Revised:2020-11-11 Online:2022-03-18 Published:2022-03-22

Abstract:

In the inheritance of shadow play culture, due to the aging of the audience and the discontinuity of inheritance, the shadow play culture is gradually facing decline. Real-time matching of shadow play movements based on Kinect can inject new vitality into traditional shadow play culture. According to the characteristics of shadow play, a joint point shadow play model is constructed, and the static digitization of shadow play is realized. The human body depth image is obtained based on Kinect, and the human skeleton point coordinates are obtained through segmentation mask and machine learning to generate the human skeleton.Bone binding and weight setting are performed on the joint point shadow play model, and the human body data collected by Kinect are bound to the shadow play. Through the traversal of the skeleton tree and the action matching algorithm, the coordinate mapping is completed to realize the real-time matching between the human body and the shadow play action, thereby completing the digitalization of the dynamic performance of the shadow play.

Key words: joint point shadow model, Kinect, depth data collection, motion capture, real-time motion matching

CLC Number: