An algorithm of texture mapping based on boundary adjustment and image blending

In this paper, A new texture mapping algorithm for 3D model is proposed. The proposed algorithm starts with a 3D model and several color images of this model. First, we match pixels of each image with the 3D model. Then the boundaries between images are determined by minimizing a boundary-based cost function. In this paper, two different cost functions are proposed. Finally, an image blending method is proposed to smooth the color around the boundaries. The proposed algorithm is robust to the illuminant difference among images and can generate texture map with much less texture discontinuity. Experimental results demonstrate the effectiveness of the proposed algorithm.