航空遥感影像正射纠正集群并行算法  

Cluster Computing Parallel Ortho-Rectification Algorithm for Aerial Images

作　　者：刘航冶[1,2] 张永生[1] 范永弘[1] 

机构地区：[1]信息工程大学测绘学院,河南郑州450052 [2]东华理工大学江西省数字国土重点实验室,江西抚州344000

出　　处：《测绘科学技术学报》2010年第6期424-428,共5页Journal of Geomatics Science and Technology

基　　金：国家863计划资助项目(2009AAZ218);江西省数字国土重点实验室开放基金资助项目(DLLJ201001)

摘　　要：正射纠正是生成正射影像的关键步骤。随着航空遥感影像数据量的急剧增加,正射纠正的处理速度问题越来越受到关注。并行处理是解决正射纠正速度问题的有效途径。在对遥感影像正射纠正算法进行分析的基础上,重点对并行正射纠正算法中的数据划分,负载平衡等关键问题进行深入研究,并提出一种适合遥感影像快速正射纠正的负载平衡并行算法。在同构集群环境下的实验结果表明,该算法能够实现集群处理节点基本的负载平衡,达到良好的并行加速比,适合于航空遥感影像的快速正射纠正。Ortho-rectification is one of the most important processing steps in generating ortho-images. As the data of remotely sensed image become larger and larger, people pay more attention to the computational speed of ortho-rectification. Parallel processing technology is an effective way to accelerate the processing speed of image ortho-rectification. Based on the analysis of ortho-rectification algorithm, data partitioning and load bal- ancing of parallel ortho-rectification were studied in deep, and a new parallel ortho-rectification algorithm for massive aerial images was given. It is proved by experiment, which was taken under homogeneous cluster environment, that this algorithm can achieve basic load balancing among processing nodes and good speed-up ratio, and is suitable for rapid processing of massive image.

关 键 词：集群并行计算 正射纠正 负载平衡 共线方程 数据划分 

分 类 号：P237[天文地球—摄影测量与遥感] TP751[天文地球—测绘科学与技术]