Enhancing Texture Classification through GLCM and LBP Feature Extraction with Modified PCA Dimensionality Reduction

Texture identification and classification is a critical function in image analysis, which has numerous applications in a variety of areas such as clinical imaging, pattern detection, and modern quality control. Experts in these fields extract valuable information from images that would otherwise be difficult to recognise by classifying textures. This information can then be used to make informed decisions, diagnose medical conditions, and optimise manufacturing processes. Therefore, the ability to accurately classify textures is a necessary skill for professionals in a wide range of domains. As a result, this paper makes use of the Grey Level Co-occurrence Matrix (GLCM) and Local Binary Pattern (LBP) techniques to extract relevant features for accurate texture classification from two different datasets, TC12 and KTH-TIPS. The modified principal component analysis (MPCA) approach is used to minimise the dimensionality of features. Dimensionality reduction is frequently utilised to reduce the complexity of the proposed approach, thereby improving performance by reducing the complexity of the approaches and making data more easily visualised.