Keypoint Reduction for Smart Image Retrieval

Content-based image retrieval (CBIR) is an image retrieval problem with image-content query. This problem has been investigated in many applications, such as, human identification, information embedding into realworld objects, life-log, and so on. Through many researches on CBIR, local image features, such as SIFT, SURF, and LBP, are proved to be effective for fast and occlusion-robust image retrieval. In CBIR using local features, not all features are necessary for image retrieval. That is, distinctive features have stronger discrimination power than commonly observed features. Also, some local features are fragile against observation distortions. This paper presents an importance measure representing both robustness and distinctiveness of a local feature based on diverse density. According to this measure, we can reduce the number of local features related to each database entry. Through experiments, we confirmed that the accuracy and the speed of image retrieval are improved by reducing the number of local features for dataset indices.