In computer vision, image retrieval remained a significant problem and recent resurgent of image retrieval also relies on other\npost processing methods to improve the accuracy instead of solely relying on good feature representation. Our method addressed\nthe shape retrieval of binary images. This paper proposes a new integration scheme to best utilize feature representation along\nwith contextual information. For feature representation we used articulation invariant representation; dynamic programming\nis then utilized for better shape matching followed by manifold learning based post processing modified mutual kNN graph to\nfurther improve the similarity score. We conducted extensive experiments on widely used MPEG-7 database of shape images by\nso-called bulls-eye score with and without normalization of modified mutual kNN graph which clearly indicates the importance of\nnormalization. Finally, our method demonstrated better results compared to other methods.We also computed the computational\ntime with another graph transduction method which clearly shows that our method is computationally very fast. Furthermore,\nto show consistency of post processing method, we also performed experiments on challenging ORL and YALE face datasets and\nimproved baseline results.
Loading....