Abstract:Texture plays an increasingly important role in computer vision. It has found wide application in remote sensing, medical diagnosis, quality control, food inspection and so forth. Research on texture started in the 1970s. The resurgence of research interest and resulting techniques in artificial neural networks gives rise to a new paradigm for texture analysis. The paper presents an application of a neural network architecture along with its training algorithm-the generating-shrinking algorithm-to texture classification in comparison with the error backpropagation algorithm and the conventional K-nearest neighbour rule. The texture feature sets considered in the paper include the statistical geometrical features and features derived from the two-dimensional discrete Fourier transform via rings and wedges

Neural Networks and Texture Classification