Research on Hand Recognition Method Based on Markov Random Field

Now, there is a problem to satisfy the demand of flexible manufacturing systems. Human-computer interaction intelligent assembly technology is more and more attention. Gesture recognition and action understanding is the basis of human-computer interaction assembly. For this reason, the research on hand gesture recognition is carried out. Firstly, the basic according to the features of the hand skin color separation method is described in this paper. The principle of YCrCb color space and the conversion between RGB color space is illustrated. It is described how YCrCb three-channel method eliminates the influence of brightness on color, and how to identify skin color and other colors. Secondly, it is explained that the principle of Markov random field. Markov random field's application in image segmentation problem is shown. Iterative algorithm process based on Markov random field is introduced. Finally, it is described how to combine the two methods, the skin color recognition and iterative algorithm. They are combined with a small area noise removal method to achieve a better image of the hand segmentation. And segmentation results and conclusions are shown.