Real-time gesture recognition based on "Bighand-Smallhand"
基于“大小手”的徒手手势实时识别

With the development of human-computer interaction technology, real-time gesture recognition based on bare hands attracts more and more attention. Most methods based on bare hands are not real-time which can only deal with data offline. A new scheme of feature extraction based on Bighand-Smallhand was proposed. The hands were divided into Bighand and Smallhand, and the overlap of both hands was treated as one hand. Totally, 17 words were tested with this method. The results indicate that the new scheme is real-time with the discrimination of 94.1%.