Texture Segmentation : Different Methods

69 Abstract—Image Segmentation is an important pixel base measurement of image processing, which often has a large impact on quantitative image analysis results. The texture is most important attribute in many image analysis or computer vision applications. The procedures developed for texture problem can be subdivided into four categories: structural approach, statistical approach, model based approach and filter based approach. Different definitions of texture are described, but more importance is given to filter based methods. Such as Fourier transform, Gabor, Thresholding, Histogram and wavelet transforms. These filters are used to VisTex images and Brodatz Textures Database. The main objective of this paper is to study different methods for texture segmentation.