Texture Image Segmentation Using Gray-Level Co-occurrence Matrix and K-Means Clustering with MATLAB Implementation

This article presents a texture image segmentation method combining Gray-Level Co-occurrence Matrix (GLCM) analysis with K-Means clustering, along with its complete MATLAB implementation. We begin by explaining the fundamental principles and algorithms behind GLCM feature extraction and K-Means clustering, providing mathematical foundations and computational approaches to ensure clear understanding of the implementation process. The discussion covers texture image segmentation background and practical applications, while analyzing the strengths and limitations of existing methodologies. Our implementation methodology is detailed through sequential stages: image preprocessing using MATLAB's image processing toolbox functions, GLCM-based feature extraction implementing statistical texture descriptors (contrast, correlation, energy, homogeneity), and clustering segmentation utilizing MATLAB's kmeans function with optimal centroid initialization. Finally, experimental results and comprehensive analysis demonstrate the method's effectiveness and advantages through quantitative evaluations and visual comparisons, including performance metrics calculation and comparative studies with alternative segmentation approaches.