Model-Based Reinforcement of Kinect Depth Data for Human Motion Capture Applications

Sensors

Luis Vicente Calderita²  Juan Pedro Bandera¹  Pablo Bustos² 

[1] Department of Electronic Technology, University of Málaga, Campus de Teatinos, Málaga 29071, Spain; E-Mail:;Polythecnic School of Cáceres, University of Extremadura, Avd. de la Universidad, Cáceres 10003, Spain; E-Mail:

关键词: human motion capture;    sensor;    RGB-D sensors;    range camera;    pose analysis;   

DOI  :  10.3390/s130708835

来源: mdpi

【 摘 要 】

Motion capture systems have recently experienced a strong evolution. New cheap depth sensors and open source frameworks, such as OpenNI, allow for perceiving human motion on-line without using invasive systems. However, these proposals do not evaluate the validity of the obtained poses. This paper addresses this issue using a model-based pose generator to complement the OpenNI human tracker. The proposed system enforces kinematics constraints, eliminates odd poses and filters sensor noise, while learning the real dimensions of the performer's body. The system is composed by a PrimeSense sensor, an OpenNI tracker and a kinematics-based filter and has been extensively tested. Experiments show that the proposed system improves pure OpenNI results at a very low computational cost.

【 授权许可】

CC BY   
© 2013 by the authors; licensee MDPI, Basel, Switzerland.