Chinese Journal of Computers   Full Text
Title	Mesh Optimization Based Image Texture Replacement
Authors	GUO Yan-Wen1) SUN Han-Qiu2) PENG Qun-Sheng3) WU Gang-Shan1)
Address	1)(National Laboratory for Novel Software Technology, Nanjing University, Nanjing 210093) 2)(Department of Computer Science and Engineering, Chinese University of Hong Kong, Hong Kong) 3)(State Key Laboratory of CAD&CG, Zhejiang University, Hangzhou 310027)
Year	2007
Issue	No.9(1580—1587)
Abstract &   Background	Abstract This paper presents a new method for replacing the existing textures in the interest region of an image with the new ones, without knowing the accurate underlying geometry and lighting conditions. A stretch-based parametrization incorporating the recovered normal information is deduced to imitate the texture distortion induced by the underlying geometry as well as the perspective distortion. The luminance of the original image is simultaneously preserved through color transfer on YCbCr color space. The method is independent on the replaced textures, however only correlates with the original image space. So after processing the input image only once, any new textures can be replaced efficiently without re-execution. By means of the parametrization, the replaced effects with progressively-variant texton scales and texture discontinuity in self-occlusion regions can be produced conveniently and realistically. Experimental results demonstrate satisfactory texture replacing results. keywords computer graphics; texture replacement; parametrization background Texture replacement or retexturing is the process of replacing the current textures on the specified region of image/video, with the same texture distortion and shading effects consistent to the unknown underlying geometry and lighting conditions. Retexturing on image and video resources has wide applications in digital entertainment, virtual exhibition, art and industry design. Existing methods on texture replacement either falls into texture synthesis based techniques with low efficiency or is only suitable to regular or near regular textures. This paper presents a texture mapping based image texture replacement method. A stretch-based parameterization scheme incorporating the recovered normal information is deduced to imitate texture distortion of the interest region. Tur approach is independent of the replaced textures, only correlating with the original image space. So after processing the input image only once, any new texture can be efficiently replaced without re-execution. Based on mesh parameterization framework, two special effects, e.g. replaced effects with progressively-variant texton scales and texture discontinuities in self-occlusion image regions, are imitated realistically. Experimental results show the satisfactory performance. This project is sponsored jointly by the Natural Science Foundation of Jiangsu Province (No.BK2007571), National Natural Science Foundation of China (Nos.60533080, 60503058, 60603086) and the CUHK Direct Research Grant (No.2050349).