Binary Genetic Programming Based Texture Segmentation

Introduction The texture is " A measure of the variation of the intensity of a surface, quantifying properties such as smoothness, coarseness, and regularity. " Textures are homogenous visual patterns in a scene such as grass, cloud, wood, and sand. Although the stated definition is an acceptable one, Zucker and Kant described the textures as: "Texture is an apparently paradoxical notion. On the one hand, it is commonly used in the early processing of visual information, especially for practical classification purposes. On the other hand, no one has succeeded in producing a commonly accepted definition of texture". Textures have spatial continuity. Gabor wavelets optimize the theoretical limit of the joint resolution between space and frequency domain. They don't have zero mean. It's used to represent the time-varying signal in terms of functions that are localized in frequency and time. The texture is essential for vision systems such as visual inspection, medical image analysis, and remote sensing. It's not yet understood how a biological process of vision happens to mimic it. In texture analysis, we have a classification and segmentation. The process of texture classification is to assign an unknown texture to known texture class while texture segmentation is identifying the region in an image based on textural differences. The thesis statement is accuracy improvement of texture segmentation using genetic programming.