An algorithm based on rough-set theory for color image segmentation

In view of the over- and under-segmentation problems existed in the conventional image segmentation based on rough-set theory, an novel color image segmentation approach based on Rough-Set theory is presented in this paper. Firstly, the new distance has been defined by using the vector angle and Euclidean distance. And then according to the new distance, the space binary matrixes that represent the similar color sphere and the Histon of each color component are calculated. Finally, the color image segmentation is implemented by selection of threshold values and region merging through introducing a histogram based on roughness. The analysis of experimental results show that the proposed approach yields better segmentation which is more intuitive to human vision compare with the conventional image segmentation based on rough-set theory.