A Novel Image Retrieval Algorithm Based on Affinity Propagation

摘要

A novel image retrieval algorithm based on affinity propagation is proposed in this paper, aiming at the disadvantages of k-means, such as choosing the initial exemplars randomly, being weak at speed, low efficiency and so on. Firstly, we extract the integrated features based on color, shape and texture from each image in the image database, and calculate the Euclidean distance pairs of image features, then establish a similarity matrix. Secondly, affinity propagation clustering algorithm clusters the image in the image database and creates the index. At last, according to the similarity criterion, similarity retrieval in the cluster is carried out. The experiment results show that, compared with k-means, the speed of affinity propagation is faster, precision and recall are more accurate , without assigning the initial exemplars. All of properties of image retrieval are also improved