Title :
Design of a Kinect Sensor Based Posture Recognition System
Author :
Chia Wen Chang ; Min Da Nian ; Yi Fang Chen ; Ching Ho Chi ; Chin Wang Tao
Author_Institution :
Dept. of Inf. & Telecommun. Eng., Ming Chuan Univ., Taoyuan, Taiwan
Abstract :
The posture recognition system based on Kinect sensor system is developed in this paper. The proposed posture recognition system contains semaphore signals, Kinect 3D sensor system, and self-organizing maps algorithm (SOM). The human-robot interaction is an intuitive and easy operation by posture recognition. However, there are not uniform postures to be a control action criterion. The flag semaphores are adopted in this paper. Then, information of skeleton point is collected by Kinect sensor. The SOM is utilized to analysis the posture information and to recognize the semaphore signal. In experiments, it notes that the posture recognition system can be accomplished to recognize the related semaphore signal.
Keywords :
human-robot interaction; image sensors; pose estimation; self-organising feature maps; Kinect 3D sensor system; Kinect sensor based posture recognition system; SOM; flag semaphores; human-robot interaction; posture recognition; self-organizing maps algorithm; semaphore signal recognition; Human-robot interaction; Image recognition; Robot sensing systems; Self-organizing feature maps; Skeleton; Testing; Three-dimensional displays; Image Recognition; Kinect; SOM;
Conference_Titel :
Intelligent Information Hiding and Multimedia Signal Processing (IIH-MSP), 2014 Tenth International Conference on
Conference_Location :
Kitakyushu
Print_ISBN :
978-1-4799-5389-9
DOI :
10.1109/IIH-MSP.2014.216
