Using GrCC for Color Image Segmentation Based on the Combination of Color and Texture

Color image segmentation has been a significant and challenging topic in the field of digital image processing. Due to the complexity of color images, the results of traditional segmentation based on granular computing clustering (GrCC) are often undesirable. In this paper, a new improvement approach based on granular computing (GrC) for color image segmentation is proposed. First, to increase the discriminability of pixels, a simple but effective filtering method is proposed. Then, to increase the discriminability of the content of an image, Gabor filter is used to analyze the texture information of the image. Thus, combining color and texture information, we use GrCC to process pixel clustering. Moreover, to obtain the segmentation result, an image is reconstructed by pixel cluster information. Finally, to evaluate the segmentation method objectively, the results of the proposed segmentation method are compared with the ground truth images. Extensive experiments performed on Microsoft Research (MSR) image data base have been conducted to validate the proposed method.