A Human Body Posture Identification Algorithm Based on Kinect

DOI

10.2991/lemcs-15.2015.1How to use a DOI?

Keywords

Posture identification; Mean Hausdorff Distance; Angle measurement; Kinect; Human-computer interaction

Abstract

This study was performed to improve the computer identify of the human body posture more accurately during human-computer interaction. In order to get a better algorithm, the author applies Kinect for Xbox device, Hausdorff Distance theory and Joint Angle measurement method to identify the human body posture. According to the actual situation, the author modifies the Hausdorff distance to mean Hausdorff distance; and sets one joint point as the reference point to improve the stability of angle measurement system. During the identification process, the Kinect gets the position of the human body joint points, and calculates the Euclidean distance between two joints, then uses cosine theorem to calculate the angle of two ligatures which are connecting three joints to define the body posture. The experimental results show that this algorithm can measure in time joint angles, and identify the body posture accurately. The study has an important reference meaning in human computer interaction.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Tong Wang
AU  - Xiaoliang Liu
AU  - Meng Wang
AU  - Tongyao Wang
PY  - 2015/07
DA  - 2015/07
TI  - A Human Body Posture Identification Algorithm Based on Kinect
BT  - Proceedings of the International Conference on Logistics, Engineering, Management and Computer Science
PB  - Atlantis Press
SP  - 1
EP  - 5
SN  - 1951-6851
UR  - https://doi.org/10.2991/lemcs-15.2015.1
DO  - 10.2991/lemcs-15.2015.1
ID  - Wang2015/07
ER  -