A New Texture Synthesis Method And A Fast Colorizing Method

Texture synthesis is one of the hottest areas in Computer Graphics, Computer Vision and Image Processing fields. In this paper, we present a new algorithm on TSfS(Texture Synthesis from Sample). Colorizing grayscale images is a useful area in Computer Vision field, We extend a texture synthesis algorithm to colorizing grayscale images field, present a fast colorizing grayscale images algorithm.Firstly, we introduce the development in the area of texture synthesis, introduce some classic algorithm on TSfS in detail. The existing method can not do well on the kind of texture that contains obvious character object, we present a method for this kind of texture. Our method detaches character object of texture from background of texture using graphcut and synthesis background of output texture using background of input texture, then put the character object on the background to get the result. Our method assures the integrality of the character object, get satisfying result.Secondly, we introduce another area in Computer Vision field: colorizing grayscale image. We introduce the development and some classic algorithm of this area. Then we present a fast algorithm. Our algorithm is inspired by a texture synthesis algorithm. In the process of the colorizing, we introduce a new searching algorithm, reduce the cost of matching pixel searching. Our algorithm can get approximate result and increase the speed of colorizing.Our texture synthesis algorithm have great potential in many applications, especially in image painting, computer animation, image compressing and so on. Our colorizing algorithm have great application in image editing and video editing.