Different Color Space Research Of Skin Color Detection Algorithm

As the development of image processing technology and network technology, in recent years the detection, identification and regional segmentation technology research of digital image color of skin is very active in science field, which achieved significant achievements. Skin detection technology is widely used in face recognition and tracking, digital image processing and the filter, characters search in database or the Internet and the medical diagnosis, and it has remarkable theoretical significance and application value.Skin detection technology based on color space has the advantages of directness, quickness, simplification, not to change from the impact of the viewpoint and shape, which has been valued in recent years by all kinds of scientific researchers, and has important theoretical and practical application value. This paper chiefly study different color space’s skin color detection algorithm.The choice of the color space and skin color modeling method is the key issue ofskin color detection algorithm. This paper comprehensively reviews and analyzes the common color space in skin color detection technology, explores clustering features of the different color space, and then studies detailedly study the method of the image preprocessing and image segmentation, static skin color model. Establishment and analysis of RGB,HSV,YCbCr three kinds of color space skin model, combining three kinds of skin color model space, comprehensively established RGB-H-CbCr new color space skin model. In order to adapt to the situation of complex light skin detection, this paper established a kind of improved YCbCr color space simple gauss skin model.Finally, the experimental simulation stage is developed through two experiments, is respectively based on the newly constructed RGB-H-CbCr hybrid skin color model skin detection system design experiment and an improved YCbCr color space simple gauss skin model experiment. For RGB-H-CbCr hybrid skin model and skin detection system design experiment, using this design method, respectively from the static and dynamic aspects of the design of the system for the experiment and simulation, through the analysis of experimental results, played a better check effect. Improved YCbCr color space simple gauss skin model have better detection rate in the light of the complex situation skin detection. From the analysis of experimental results this paper basically reached this topic’s research purposes.