Texture and Color based Image Retrieval using Gray Level Co-Occurrence Matrix

Abstract: Content Based Image Retrieval (CBIR) allows us to retrieve most similar image/images from the database based on the visual content of the image. The major steps being feature extraction and computing similarity. In this paper we extracted 17 color and texture features of image. To extract color feature we use F-Norm and to extract texture feature we use statistical measures (mean, standard deviation) and GLCM that is being applied at four different angles. The features thus extracted are then subjected to similarity measures using Euclidean distance. This technique showed a significant improvement in retrieval performance when compared to other methods such as Weighted Standard Deviation (WSD), Gradient operation using Sobel operator. Keywords: CBIR, Gray Level Co occurrence Matrix (GLCM), F-Norm.