A novel image clustering algorithm on peer-to-peer network

In order to decrease transmission data on P2P network and improve image clustering efficiency, an image clustering method is presented in the paper. We define the concepts of peer, class, group and overlay, which have their representation or signature, and describe the architecture of peer-to-peer network. In the method, the feature space is divided into units into which the feature vectors may be mapped, thus images can be represents by a simpler manner according to the partition units. We propose class similarity and group similarity algorithms, and give the peer clustering strategy. we use these methods to build the overlay network on which peers in one overlay are similar. Finally, the experiment results show that the proposed image clustering algorithm can reduce transmission content in the network while increase the performance.