Local search optimized hashing for fast image copy detection

Recently, researches on content based image copy detection mainly focus on robust feature extraction. However, due to the exponential growth of online images, it is time-consuming and unscalable to search among large scale images. Although many hashing methods has been proposed to improve the efficiency of image copy detection, they confront semantic loss issue. In this paper, we propose a new hashing based method for fast image copy detection. It first generates compact fingerprint which combine the influence of both the neighborhood structure of feature data and mapping error to prevent huge semantic loss during the process of hashing. Then optimize the solution through Local Search to further decrease semantic loss. Experimental results show that our approach significantly outperforms state-of-art methods.