Similarity based retrieval of image database: using dynamic clustering

In this paper, a ranking algorithm using dynamic clustering for content-based image retrieval (CBIR) is proposed. In conventional CBIR systems, it is often observed that visually dissimilar images to the query image are located at high ranking. To remedy this problem, similarity relationship of retrieved results are utilized via dynamic clustering. In the first step of this method, images are retrieved using visual feature such as color histogram, etc. Next, the retrieved images are analyzed using a HACM (hierarchical agglomerative clustering method) and the ranking of results are adjusted according to distance from a cluster representative to a query. This paper shows the experimental results based on MPEG-7 color test images. According to the experiments, the proposed method achieves more than 10% improvements of retrieval effectiveness in ANMRR (average normalized modified retrieval rank) performance measure.