Texture image segmentation technique is a key technique in Digital Image Processing. As a basis of recognizing, analysing and understanding texture image, Image segmentation exact the interested characteristic including edge and region, etc. Therefore, the research of texture image segmentation is going deeply more and more, and became a research hot in image processing.This paper has discussed the problem of detecting edges in cellar textures. A general detection procedure was proposed. The procedure involved applying an edge-sensitive operator to the texture, thresholding the results of the edge operator, and finally computing"peaks"from the above-threshold points. The paper concentrated on the thresholding process and developed a minimum-error thresholding process based on an analysis of the edge operator. The thresholding process assumed that it was normally distributed at edges and at interior points. All the work in this paper is of some value to research and applications of image edge detection. |