Applications Of Quaternion In Skeleton Skin Technology

In human daily life, there is a lot of colorful information. This makes image processing being an active field in computer applications. Digital image processing has been widely applied in advertisment, film and television making, and other business activities.Digital image matting is a kind of image processing technology that extracts a foreground element of arbitrary shape from a natural image. Blue screen matting and differentiation matting, developed in the early years, generally impose strict requirements on the background of the image. It makes the matting technology significantly limited and cannot meet the demands of applications, a variety of image matting algorithms were proposed in order to obtain perfect matting effect. This includes bayesian matting, robust matting and Poisson matting etc..In this paper, an in-depth study is performed on some typical digital image matting algorithms. Concretely, we compare the running efficiency, the matting quality, and limitations of Bayes matting, Poisson matting and Robust matting, through a great number of experiments, we can see that:(1) The size and the elaborate degree of the input picture decide the The rate of matting.(2) When the same input picture and the same trimap, The quality of Robust matting are better than Bayes matting and Poisson matting.(3) In the Bayes matting, you have to cost a lot of time of human-computer interaction to get better matting effect.