Research on parallel unsupervised classification performance of remote sensing image based on MPI

The rapid processing of mass remote sensing data put challenges on computer's processing capability. Through parallel programming environment based on message passing, parallel K-means unsupervised classification of remote sensing image with different sizes we performed in parallel environment with different computers number. The speedup and efficiency of parallel computation as well as effect of message communication on parallel unsupervised classification were analyzed. The results show that the classification speed of mass amount data parallel remote sensing image unsupervised classification has been greatly improved and the parallel unsupervised classification has effect on parallel efficiency. In the parallel programming, message communication should be minimized as possible and messages should be merged to improve computation efficiency. Rational task allocation and communication can improve performance of parallel computing. (C) 2011 Elsevier GmbH. All rights reserved.