Texture Analysis by Gray Level Homogeneity in Local Regions

In this paper, we propose a method that helps us analyze the texture of an image through the study of the appearance and recognition of the shapes of the regions of the image by studying the homogeneity of the gray level among the local regions. This principle, based on the coding of the pixel according to its adjacent neighbors, allows us to describe the nature of the texture and measure the degree of local gray level homogeneity. This algorithm gives us an idea of the change in contrast and the flow of gray levels between regions, and it also detects the areas of interest in the image. This method is interested in studying the homogeneity of gray levels like other methods such as the local contrast descriptor (LCD) and Co-occurrence Matrix, so that it is easy and effective to discover the borders of regions in images, improve their weak contours, and reduce noise.