Particular Video Object Extraction Technology Research

In the modern society, with the development of the TV technology in the digital and the multimedia way, matting and compositing becomes the new focus in 21 century. As the key of visual effect production, the matting technology in video will be a good future. Despite having enjoyed wide usage for many years, traditional matting and compositing have limitations. Traditional matting methods either require special setups or cannot handle objects with complex silhouettes. To overcome these disadvantages above and expand the scope of blue matting technology, the author researches how to extract object in constant color background, generalized constant color background and natural background in video in image manage way.In this paper the author firstly introduces the matting technology and discusses the methods of classifying the matting algorithm in the paper. In the constant color background condition, a matting method based on threshold is presented. The author uses improved Euclid Distance as measurement to extract the object and tests the method in several color spaces. Then, the author presents a blue matting method based on edge detection which uses several classic image segmentation algorithms together. The tests prove these two methods are effective under the constant color background condition and can be used to extract objects in generalized constant color background. However, the results under generalized constant color background are not so accurate. To heighten the effects, the author introduces the α estimation arithmetic in the generalized constant color background and natural background condition. Lots of tests prove that the results are higher quality than those using traditional image partition arithmetic. Besides, the method expands the blue matting technology and satisfies the broadcast level quality need.