可扩展的迭代自组织分析并行处理算法An improved parallel remote sensing ISODATA algorithm
夏辉宇;
摘要(Abstract)：
随着遥感影像数据量的增加,传统非监督分类迭代自组织分析(ISODATA)算法的运算将十分耗时,应用并行计算技术能够有效解决该性能瓶颈。针对现有基于并行计算模型MapReduce的遥感迭代自组织分析并行算法存在的局限性,提出一种可扩展的基于MapReduce的迭代自组织分析并行处理算法。该算法通过其包含的全局子采样算法、聚类中心点集合过滤算法以及聚类映射算法,有效克服了现有并行算法中存在的不足。实验结果表明,在同等规模遥感计算中,该算法效率高于现有并行处理算法,具有良好的加速比,且在处理更大的影像块时具有更高的精度。
关键词(KeyWords)： 迭代自组织分析算法;ISODATA;MapReduce模型;并行计算;遥感影像;非监督分类
基金项目(Foundation):
作者(Authors): 夏辉宇;
DOI: 10.16251/j.cnki.1009-2307.2016.08.002
参考文献(References)：
- [1]CHEN Z,CHEN N,YANG C.Cloud computing enabled web processing service for earth observation data processing[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2012,5(6):1637-1649.
- [2]BALL G,HALL D.ISODATA,a novel method of data analysis and pattern classification[R].[S.l.]:Stanford Research Institute Menlo Park CA,TIC Document,1965.
- [3]AHMAD A,SUFAHANI S.Analysis of Landsat 5TM data of Malaysian land covers using ISODATA clustering technique[C]//2012IEEE Asia-Pacific Conference on Applied Electromagnetics(APACE).USA:IEEE,2012:92-97.
- [4]AL-AHMADI F,HAMES A.Comparison of four classification methods to extract land use and land cover from raw satellite images for some remote Arid Areas,Kingdom of Saudi Arabia[J].Journal of King Abdulaziz University-Earth Sciences,2009,20(1):167-191.
- [5]JOYCE K,BELLISS S,SAMSONOV S.A review of the status of satellite remote sensing and image processing techniques for mapping natural hazards and disasters[J].Progress in Physical Geography,2009,33(2):183-207.
- [6]MEMARSADEGHI N,MOUNT D,NETANYAHU N.A fast implementation of the ISODATA clustering algorithm[J].International Journal of Computational Geometry&Applications,2007,17(1):71-103.
- [7]刘向东,刘奎,胡飞翔,等.基于MapReduce的并行聚类算法设计与实现[J].计算机应用与软件,2014(11):251-256.
- [8]DHODHI M,SAGHRI J.D-ISODATA:a distributed algorithm for unsupervised classification of remotely sensed data on network of workstations[J].Journal of Parallel and Distributed Computing,1999,59(2):280-301.
- [9]PLAZA A.Parallel techniques for information extraction from hyperspectral imagery using heterogeneous networks of workstations[J].Journal of Parallel and Distributed Computing,2008,68(1):93-111.
- [10]YE F,SHI X.Parallelizing ISODATA Algorithm for unsupervised image classification on GPU[J].Modern Accelerator Technologies for Geographic Information Science,2013:145-156.
- [11]LV Z,HU Y,ZHONG H.Parallel K-means clustering of remote sensing images based on MapReduce[C]//Proceedings of the 2010International Conference on Web Information Systems and Mining.Berlin:SpringerVerlag,2010:162-170.
- [12]徐昌荣,王聪颖,袁秀华.基于并行编程计算模型的索贝尔滤波技术[J].测绘科学,2014,39(10):85-88.
- [13]ALMEER M.Cloud Hadoop map reduce for remote sensing image analysis[J].Journal of Emerging Trends in Computing and Information Sciences,2012,3(4):637-644.
- [14]LI B,ZHAO H,LV Z.Parallel ISODATA clustering of remote sensing images based on MapReduce[C]//2010International Conference on Cyber-Enabled Distributed Computing and Knowledge Discovery(CyberC).USA:IEEE,2010:380-383.
- [15]BU Y,HOWE B,BALAZINSKA M.HaLoop:efficient iterative data processing on large clusters[J].Proceedings of the VLDB Endowment,2010,3(1/2):1-2.
- [16]SRIRAMA S,BATRASHEV O,JAKOVITS P,et al.Scalability of parallel scientific applications on the cloud[J].Scientific Programming,2011,19(23):91-105.
- [17]HORE P,HALL L,GOLDGOF D.A scalable framework for cluster ensembles[J].Pattern Recognition,2009,42(5):676-688.