[1]岳钦崟,李 虎.基于 MPI 并行遥感典型地类提取应用[J].计算机技术与发展,2020,30(07):104-108.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 023]
　YUE Qin-yin,LI Hu.Application of MPI-based Parallel Remote Sensing for Typical Ground Classification Extraction[J].,2020,30(07):104-108.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 023]
点击复制

基于 MPI 并行遥感典型地类提取应用()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
30
期数:
2020年07期
页码:
104-108
栏目:
应用开发研究
出版日期:
2020-07-10

文章信息/Info

Title:
Application of MPI-based Parallel Remote Sensing for Typical Ground Classification Extraction
文章编号:
1673-629X(2020)07-0104-05
作者:
岳钦崟; 李 虎
中国科学院国家空间科学中心,北京 100190
Author(s):
YUE Qin-yin; LI Hu
National Space Science Center,Chinese Academy of Sciences,Beijing 100190,China
关键词:
MPI 协议; 并行任务处理; 遥感图像; 异构编程; 地物识别
Keywords:
MPI protocol; task concurrent processing; remote sensing image; heterogeneous programming; feature interpretation
分类号:
TP277
DOI:
10. 3969 / j. issn. 1673-629X. 2020. 07. 023
摘要:
遥感图像地物识别是一项耗时耗资源的处理过程,实际应用中需实现对典型地类的快速识别才能展现遥感卫星具有高频次重访、高时效观测、快速机动观测的特点和优势。 该过程涉及海量数据和多种地类特征识别算法,这些算法实现方式各异,与图像数据产品生产和业务系统处于差异化的平台环境下,这给遥感图像数据产品生产和业务系统构建带来了一定的困难。文中提出了一种基于 MPI 高性能并行消息处理模型的典型地类特征高性能处理模型和框架,能有效地支持业务系统和地类特征识别算法的异构开发环境,并提供多节点多进程的高性能并行任务处理,具有支持异构编程、海量数据高性能处理、跨平台可移植性等特点。 该模型框架为自然灾害监测、气候变化研究、农林资源监测、地震环境调查、陆地植被调查、森林及水资源调查等工作的典型地物识别、产品生产以及业务系统运行提供技术支撑和方法指导。
Abstract:
Feature interpretation in remote sensing is a time-consuming and resource-consuming procedure. In order to show the characteristics and advantages of remote sensing satellites with high frequency re-visit,high time-efficiency observation and fast maneuverable observation,it is necessary to realize rapid recognition of typical terrain types in practical application. This process involves massive data and many kinds of land feature recognition algorithms which are implemented in different ways. They are developed based on different libraries or different platforms environment,and integrated with the production and business system of image data products. This brings some difficulties to the production and business system construction of remote sensing image data products. We present a high performance processing model and framework of typical geographic features based on MPI model, which can effectively support heterogeneous development environment of business system and geographic feature recognition algorithm and provide high performance parallel task processing of multi-node and multi-process,with heterogeneous programming supporting, massive data high-performance processing,cross- platform portability and other characteristics. The model framework provides technical support and methodological guidance for typical features identification,product production and operation of business system in natural disaster monitoring,climate change research,agricultural and forestry resources monitoring,earthquake environment survey,land vegetation survey,forest and water resources survey.
更新日期/Last Update: 2020-07-10