DisClus: A distributed clustering technique over high resolution satellite data

Abstract

This paper presents a distributed Grid-Density based Satellite data Clustering technique, DisClus, which can detect clusters of arbitrary shapes and sizes over high resolution, multi-spectral satellite datasets. Quality of the clusters is further enhanced by incorporating a partitioning based method for the reassignment of the border pixels to the most relevant clusters. Experimental results are presented to establish the superiority of the technique in terms of scale-up, speedup as well as cluster quality. © 2010 Springer.