Texture image segmentation method based on multilayer CNN

The paper presents a new texture feature extraction method called simple texel scale feature (STSF) based on the scale and orientation information of texels, and a new texture image segmentation method based on binary image processing is introduced. The scale information of texels is extracted by comparing the gray value of two pixels. The relation of the positions of these two pixels shows the frequency and orientation features of texels. Texel scale features can be extracted by using different position relations (distance and orientation). After obtaining texture feature images, we consider the texture image segmentation problem not as a pattern classification problem but several texture edge integration problems, which are simple binary value line processing problems such as hole filling, line thinning and shortening. A new kind of multilayer cellular neural network (CNN) called MLCNN is proposed, and some MLCNNs are designed for these problems