Performance evaluation of GLCM and pixel intensity matrix for skin texture analysis

SummaryTexture refers to visual patterns or spatial arrangement of pixels. Texture analysis is one of the features in image processing that is used to analyze the images captured by the imaging devices on human skin. Skin texture analysis plays a vital role in assessing the skin health and in the diagnosis of skin disorders such as allergic skin disorders, viral skin disease, bacterial skin diseases and fungal skin diseases. Texture analysis is carried out by one of the methods namely, structural, statistical, model based, transform based techniques. Statistical texture analysis depends mainly on feature extraction which may be done using GLCM (grey level co-occurrence matrix) and WDM (wavelength division multiplexing) techniques. The extracted features are used to classify texture. In this paper a study of skin texture analysis is carried out by comparing GLCM features and pixel intensity matrix parameters by experimenting with cheek and dorsal skin samples. The results obtained show that pixel intensity matrix parameters are more helpful than GLCM for analyzing the texture of skin.