Abstract

Content Based Image Retrieval systems based on shape using invariant image moments, viz., Moment Invariants\nand Zernike Moments are available in the literature. Moment Invariants and Zernike Moments are good at representing the\nshape features of an image. Therefore, an efficient and orthogonal moment based Content Based Image Retrieval system\n(CBIR) is needed. Legendre Moments is orthogonal, computationally faster, and can represent image shape features\ncompactly. CBIR system using Exact Legendre Moments and Contourlet Transform is proposed in this work. The results of\nproposed CBIR model are robust, efficient and effectively reduce the retrieval time and also improve the retrieval accuracy\nsignificantly. When compared to the other moment based CBIR systems, it is found that, the performance measures such as\nprecision and recall values are improved.