Image texture classification using gray level co-occurrence matrix and neural network
Muhammad Suzuri Hitam, Mohd Yahyal Haq Muslan, Mustafa Mat Deris, Md Yazid Mohd Saman · 2003
This paper presents the use of Grey Level Co-occurrence Matrix (GLCM) method together with multilayer fully-connected feed-forward perceptron (MLP) for image texture classification. GLCM method was employed as a features extractor technique and MLP was used as a image texture classifier. A range of Brodatz textures were employed to evaluate the proposed system. Results from various experimental investigations including general image textures, noise added images textures and rotated image textures showed that the MLP with GLCM works well as a texture classifier.