Texture Image Classification Using Support Vector Machine

Texture refers to properties that represent thesurface or structure of an object and is defined assomething consisting of mutually related elements.The main focus in this study is to do texturesegmentation and classification for texture images.Statistical features can be calculated based on thegrey level co-occurrence probabilities (GLCP)generated. The statistical features used in this studyare uniformity, contrast, and entropy. The featuresare obtained by using a combination of differentangles. For noise reduction, an appropriate movingaverage is applied to the statistical features. To postprocessthe image, support vector machines (SVM)had been proposed to do classification on theextracted features. Some kernel functions which arebeing tested are second degree polynomial, radialbasis function (RBF), exponential radial basisfunction (ERBF), sigmoid, and odd-order Bspline.RBF and ERBF achieved the best classificationaccuracy compare to other kernels used. SVM alsoautomatically helps RBF kernel to define the centresduring optimization. Brodatz texture album is used inthis study to test out the result. In the study, acombined GLCP with SVM post-processing showeda marked improvement over other classifier in termsof classification accuracy.