Image Clustering Using Mean Shift Algorithm

Clustering is an important technique of image analysis. Mean shift based clustering is a nonparametric method and can find arbitrary shape clusters in the feature space of an image. In this paper, we present a re-clustering technique based on the mean shift algorithm. The proposed method consists of two steps. First step divides the original image into homogenous image segments. Second step clusters the image segments to obtain the result of image clustering. An experiment is carried out on a remote sensing image. In the experiment, the method is compared with the traditional mean shift based image clustering. The results show that this method is better than the traditional method in classification accuracy.