Automatic Human Joint Detection Using Microsoft Kinect

Samuel Cahyawijaya, Iping Supriana Suwardi


Automatic human joint detection has been used in many application nowadays. In this paper, we propose an approach to detect full body human joint method using depth and color image. The proposed solution is divided into 3 stage, which is image preprocess stage, distance transform stage, and anthropometric constraint analysis stage. The output of our solution is a stickman model with the same pose as in the given input image. Our implementation is done by using a Microsoft Kinect RGB and depth camera with 480x640 image resolution. The performance of this solution is demonstrated on several human posture.


Microsoft Kinect, human joint, anthropometry, depth image, body measure


Dedeoglu, Y. (2004). Moving Object Detection, Tracking and Classification for Smart Video Surveillance. Ankara, Turkey: bilkent university.

Gritai, A., & Shah, M. (2006). Tracking of human body joints using anthropometry . Int. Conf. of Multimedia and Expo. Florida: University of Central Florida.

Hamissi, M., & Faez, K. (2013). Real-Time Hand Gesture Recognition Based on the Depth Map. International Journal of Electrical and Computer Engineering, 2-4.

Jain, H. P., & Subramanian, A. (2011). Real-time Upper-body Human Pose Estimation using a Depth Camera. HP Laboratories.

Kar, A. (2010). Skeletal Tracking using Microsoft Kinect. Kanpur: Methodology.

MA, B., XU, W., & WANG, S. (2013). A Robot Control System Based on Gesture Recognition. TELKOMNIKA, Vol. 1, No. 5.

NASA. (1978). Anthropometric Source Book Volume I: Anthropometry for Designers. Springfield: NASA.

Shotton, J., Fitzgibbon, A., Cook, M., Sharp, T., Finocchio, M., Moore, R., et al. (2011). Real-Time Human Pose Recognition in Parts from Single Depth Images. IEEE Conference. Microsoft Research Cambridge & Xbox Incubation.

Full Text: PDF


  • There are currently no refbacks.