Research Of The Gamut Extension Mapping Based On Wide Gamut Display

With the rapid development of a wide color gamut display device, laser show, with high saturation, high brightness and resolution, obtains more and more attention and becomes the focus of research in the display field. The coverage of the laser display color gamut, compared with the NTSC standard, is 1.8 times, and compared with the PAL standard, is more than 1.9 times. However, due to the existing standard and laser color gamut domain not matching, the color signal cannot be displayed directly in the laser display device, which will cause distorted color. Therefore, in order to demonstrate the advantages of the high saturation and brightness of the wide color gamut display fully, color space conversion and color gamut extension mapping become an indispensable steps and important part. Then, to develop applicable color gamut extension mapping algorithm is very important.With the research of the color gamut extension, it reveals other problems, such as, overlooking the color in the process of the extension, which is easy to cause the problem of the excessive rendering and the confusion of the skin color and hair color, to destruct the skin color distribution of the original image. This paper aim at the problems, mainly study the skin color protection algorithms during gamut mapping. We firstly describe the basic theory of the color gamut extension mapping, including the principles and research methods. Then followed by the instruction of existing models of skin color, contrast and compare to choose the model used in this article. On these bases, the protection algorithms of the color gamut extension mapping can be studied depth.In this paper, propose two color protection algorithms:change hue protection algorithm and based on probabilistic model protection algorithm. The change hue protection algorithm combined with linear and nonlinear extension algorithm for different hues, making the color to change smoothly and eliminating level phenomenon during the drawing process. However, this method cannot guarantee the values of the skin color area boundary remain unchanged and skin color cannot be well protected when skin color close to hair color. The based on probabilistic model protection algorithm firstly determine the skin area, when the face skin color close to hair color, we calculate the minimum of the energy to determine the color border. For the non-skin color, we extend normally; otherwise, take the smaller coefficient to extend, which ensure the color natural after extension. With the gamut extension experiments of lots of pictures show that, the above skin color protection algorithm is effective and has effective image gamut extension result, not only to eliminate the excessive rendering problem, but also to extend accurately for skin color.