Application of improved gesture recognition algorithm based-on Density Distribution Feature

Basing on the Density Distribution Feature method, present a new hand recognition method according to color space which could recognize hand in real time vision. First, establish the standard Skin-Color Model, and make the Skin-Color model adapt to different users´ skin color which is to segment the skin color accurately. Then use the density distribution feature of hand to recognize hand gestures, and the shape features of gestures could be obtained by computing the disposition of pixel points in hand areas with different distance. Experimentations show that, the improved method could detection hand in complex background, no meter if the rotation angel and proportion are large or small. Several experimental images are presented to support the validity of the method.