Identification of Similar Looking Bulk Split Grams using GLCM and CGLCM Texture Features

Content based image retrieval (CBIR) is an automated way to retrieve images based on the visual content or image features itself. Visual inspection of food type is tiresome and time consuming task. This paper presents the retrieval of similar looking bulk split gram images using Grey Level Co-occurrence Matrix (GLCM) and Color Grey Level Co-occurrence Matrix (CGLCM) texture features. Texture feature matching procedure is based on three distance measures namely, Euclidean distance, Canberra distance and City block distance. The performance of a retrieved image is measured in terms of Precision. Experimental results show that the CGLCM provides better retrieving result than GLCM.