Image Texture, Texture Features, and Image Texture Classification and Segmentation

Abstract

In this chapter, we will discuss the basic concept of image texture, texture features, and image texture classification and segmentation. These concepts will be the foundation to understand image texture models and algorithms used for image texture analysis. Once texture features are available, many classification and segmentation algorithms from traditional pattern recognition can be utilized for labeling textural classes. Image texture analysis strongly depends on the spatial relationships among gray levels of pixels. Therefore, methods for texture feature extraction are developed by looking at this spatial relationship. For example, the gray-level co-occurrence matrix (GLCM) and local binary patterns (LBP) were derived based on this spatial concept. Traditional techniques for image texture analysis, including, classification and segmentation, fall into one of the four categories: statistical, structural, model-based, and transform-based methods. The rapid advancement of deep machine learning in artificial intelligence and convolutional neural networks (CNN) has been widely used in image texture analysis. It would be essential for us to further explore image texture analysis with deep CNN.