Depth-based Static Sign Language Recognition System

Kisang Kim, Sukyung Kim, Hyung-Il Choi
2014 unpublished
In this paper, we propose the hand gesture recognition, especially static sign language gesture, using depth camera. For this system, we developed a finger extraction and gesture recognition method. In the extraction, we detect hand skeleton with Distance Transform. To extract fingertip, we use Convex Hull with hand skeleton. With hand skeleton, we could detect more accurate fingertip than contour-based method. For recognition, we create decision-tree with several features, hand center, hand
more » ... s, finger length, finger axis and arm center. Through the experimental results, we show that the performance is successful with our methods.
doi:10.14257/astl.2014.63.44 fatcat:hi3ghozy6rbunk7eu54oeg5kf4