Image resolution dependency of Local Texture Patterns in classification of color images

In this paper, experiments have been conducted to study the significance of the dots per inch resolution of the color images in texture classification. Textural features of the image are extracted using Gray-Local Texture Patterns (GLTP) operator which is an extended version of Local Texture Patterns (LTP) texture model. Contrast being another important property of images, Color-Local Contrast Variance Patterns (CLCVP) is used to extract contrast feature. However, much important information contained in the image can be revealed by joint distributions of individual features. Hence, GLTP/CLCVP is used as a textural feature extraction technique for classification of color images. The performance of the GLTP/CLCVP operator is tested for classification of images from Outex texture database. From the experimental results, it is observed that the classification accuracy is highly dependent on the dots per inch resolution of the images.