Color human slice segmentation based on texture information

This paper presents a method to segment the muscle issue in color human slice, which firstly turns the color human slice into gray image, then segment the muscle tissue based on the texture information utilizing threshold segmentation. We use the gray level co-occurrence matrix (GLCM) to characterize the texture information, and then analyze the GLCM´s parameters including the pixel pitch, the gray level and the template size to get appropriate values for the coming process. The experimental result shows that, this method can make good effect on the segmentation of muscle issue, locates the muscle tissue exactly and has higher robustness for noise points.