Robust gesture recognition based on distance distribution feature and skin-color segmentation

A new hand gesture recognition method combined color spaces and feature region of the hand is pretended in this paper, which is efficient to solve the problem that the accuracy and efficiency could not be positively related to each other in traditional hand gesture recognition methods. Use skin color model and users´ skin color to realize color region segmentation, and use color markers to exclude the interference type color pixel of the non-hand regions. The main novelty of the proposed approach is the use of distance distribution feature to recognize the hand gestures, which can overcome the instability of the hand gesture recognition caused by the changing distance and direction of user´s hand. Experimental results show that this method has good robustness and it can detect and recognize the hand gestures in varying illumination conditions, hand distance and hand angels efficiently.