基于MPI和MapReduce混合编程的高分三号多模数据分布式并行转换算法  

Distributed Parallel Conversion Algorithm of GF-3 Multi-Mode Data Based on MPI and MapReduce Hybrid Programming

作　　者：张如明 蔡剑英 王锡航 刘子立 欧阳波 ZHANG Ruming;CAI Jianying;WANG Xihang;LIU Zili;OUYANG Bo(Hulunbeir Big Data Center;Hulunbeir Smart City Development Research Institute,Hulunbeir,Inner Mongolia Autonomous Region,021000 China)

机构地区：[1]呼伦贝尔市大数据中心 [2]呼伦贝尔智慧城市发展研究院,内蒙古呼伦贝尔021000

出　　处：《科技资讯》2024年第14期34-36,共3页Science & Technology Information

基　　金：2023年呼伦贝尔市科技项目“呼伦贝尔市空间数字底座”(项目编号:GX2023004)。

摘　　要：我国高分三号形成三星陆海雷达卫星星座。为了提升PolSARpro极化SAR图像处理的开源软件在高分三号数据转换环节的处理能力,提出了基于OpenMP、MapReduce和MPI并按照PolSARpro软件的数据格式要求进行分布式并行转换处理算法,将不同成像模式下的多极化数据实现快速定标和格式转换,并且通过KingMap V10.0 GIS平台实现了算法并在实际数据中进行测试,验证了算法的高效性、可靠性、易扩展性和正确性。China's GF-3 is formed a land-sea radar satellite constellation of 3 satellites.In order to improve the processing capability of PolSARpro open source software for polarimetric SAR image in the data conversion stage of GF-3,this paper proposes a distributed parallel conversion algorithm according to the data format requirements of its software based on OpenMP,MapReduce and MPI,which can quickly calibrate and convert GF-3 mutipolarization of different imaging modes data into bin format.The algorithm is implemented on GIS platform of KingMap V10.0 and tested in real data,which verifies the efficiency,feasibility,extendibility and correctness of the algorithm.

关 键 词：高分三号 合成孔径雷达 定标 消息传递接口 分布式并行计算 

分 类 号：P237[天文地球—摄影测量与遥感]