Dual directional multi-motif XOR patterns: A new feature descriptor for image indexing and retrieval

In this paper a new feature descriptor ‘dual directional multi-motif XOR pattern (DDMMXoRP)’ is proposed for image indexing and retrieval. This method is entirely different from the existing motif representation. Here we introduced a new methodology of extracting one standard 2 × 2 grid at distance two and four 1 × 3 smart grids along dual directions which are not present in existing motif representation. This entire operation is implemented on ‘V’ color space of HSV color plane. Further, the XOR operation is performed on the transformed new motif images which are not present in the literature (local binary patterns (LBP) and motif co-occurrence matrix (MCM)). The performance of the proposed method is tested by conducting two experiments on Corel-5000 and Corel-10,000 databases. Experimental results demonstrate that it is much more efficient in terms of average retrieval precision (ARP) and average retrieval rate (ARR) than existing image feature descriptors such as, local binary patterns, motif co-occurrence matrix, multi texton histogram and other state-of-the-art techniques for image retrieval.