网格环境下批量遥感图像处理  

Processing of batch RS image in grid environment

作　　者：张旭晴[1] 陈圣波[1] 范继璋[1] 魏晓辉[2] 金晟业[1] 

机构地区：[1]吉林大学地球探测科学与技术学院,长春130026 [2]吉林大学计算机科学与技术学院,长春130026

出　　处：《世界地质》2009年第3期379-383,共5页World Geology

基　　金：国家自然科学基金项目(40471086);国家863计划项目(2006AA12Z102)联合资助

摘　　要：网格环境下遥感图像的并行处理,按照任务分解模式的不同可分为单图像多进程及单图像单进程两种任务划分方案。在对两者对比分析的基础上,针对批量遥感图像处理任务的特点,采用单图像单进程任务划分方案,在MPICH-G2模式下利用GridFTP模块解决海量遥感图像传输问题,使用Intel公司的OpenCV开源库编写繁杂图像处理代码,并基于"嫦娥1号"卫星首幅遥感影像进行环状构造边缘检测实验,以此对本文采用的批量遥感图像处理模式加以验证。According to the difference of the task separation pattern, parallel processing of batch RS image in grid environment can be divided into two programs, including single picture-multilateral process and single picturesingle process. Based on the comparative analysis, the authors adopt single picture-single process, and use GridFTP under MPICH-G2 model to solve the problem of massive RS pictures transmission aiming at the characteristics of batch RS image processing task. Using the open source library of OpenCV from Intel company to compile complicated picture handling code, and processing detection experiment of circular structure edge base on the first image of Chang＇e-1 satellite, validate the pattern of the process of batch RS image in this paper.

关 键 词：网格 MPICH-G2 OPENCV GLOBUS GRIDFTP 批量遥感图像处理 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]