Title :
Arm movement recognition for flag signaling with Kinect sensor
Author_Institution :
Dept. of Intell. Syst., Hiroshima City Univ., Hiroshima, Japan
Abstract :
This paper propose a simple recognition method for arm movement in Japanese flag semaphore. The method consists of three main process, detection of an arm position, discrimination of a semaphore form to the detected one, and identification of a Japanese Katakana character from a sequence of semaphore forms. Based on the method, an experimental system for flag signaling was implemented by Kinect sensor device and OpenNI/NITE software library. A preliminary experiment indicated that the method might be usable for beginner to learn the flag signaling. But it was also found that the method should be improved in some misrecognition.
Keywords :
gesture recognition; image motion analysis; software libraries; Japanese Katakana character; Japanese flag semaphore; Kinect sensor device; OpenNI/NITE software library; arm movement recognition; arm position detection; flag signaling; semaphore forms; Computers; Elbow; Equations; Humans; Shape; Shoulder; Virtual environments; flag semaphore; game; skill acquisition; virtual learning environment;
Conference_Titel :
Virtual Environments Human-Computer Interfaces and Measurement Systems (VECIMS), 2012 IEEE International Conference on
Conference_Location :
Tianjin
Print_ISBN :
978-1-4577-1758-1
DOI :
10.1109/VECIMS.2012.6273224