The Optimal Clustering Algorithm for Image Organization and Management in Wireless Mobile Internet

In modern multimedia applications, images are often organized and managed by web sites or local applications with geographic information like GPS. But now this kind of multimedia applications doesn´t come up in mobile communication environment. This paper proposes a method in image organization and sharing on mobile terminal whose resources are limited. We use the K-means clustering algorithm and its optimized algorithm to accomplish the image organization with geographic information. The application is based on C/S architecture under mobile environment. By comparing the optimized K-means algorithm with the original one, we get the conclusion that the optimized algorithm achieves a much better partition of the images. The demo mentioned at the end of the paper has validated the feasibility and advantages of this scheme.