Detection of hands-raising gestures using shape and edge features
ROBIO'09 Proceedings of the 2009 international conference on Robotics and biomimetics
| Hi-index | 0.03 |
Detecting human arm motion in a typical classroomenvironment is a challenging task due to the noisy andhighly dynamic background, varying light conditions, aswell as the small size and multiple number of possiblematched objects. A robust vision system that can detectevents of students' hands being raised for asking questionsis described. This system is intended to support thecollaborative demands of distributed classroom lecturingand further serve as a test case for real-time gesturerecognition vision systems. Various techniques includingtemporal and spatial segmentation, skin coloridentification, as well as shape and feature analysis areinvestigated and discussed. Limitations and problems arealso analyzed and testing results are illustrated.