The Research And Design Of Grid-based Remote Sensing Products Services

In recent years, remote sensing technology develops rapidly and is used in various fields of thenational economy, and has played a major role for promoting social progress and economic construction,national defense construction and environmental improvement. All relevant units are more or less involvedin the research of remote sensing technology and get a certain achievements. In order to fully rational useof the existing achievements resources, the effective resource sharing mechanism between the various unitswill be established to share their advantage through grid technology, the units set up earth observation gridsystem, providing cross-department, comprehensive, standardized, multi-level and multi-field networkinformation services.For Remote Sensing production services, access the existing resources for example algorithms whichare used to produce remote sensing products to earth observation grid as a mesh node to provide remotesensing products services which is essential part of earth observation grid. Remote sensing production gridis mainly used by all levels users through the3-5level production services.This paper is based on grid technology and remote sensing earth observation grid system to designremote sensing production grid node and realize a remote sensing data processing grid node prototypesystems which is accessed to earth observation grid system to share the resources, the main work are asfollows.1. Base on remote sensing image data processing and the construction tasks and goals of earthobservation grid, as well as the production and search service mode, analyze the popular grid nodearchitecture and design the grid nodes service functionality and architecture. The core objective is to study the grid technology and its mainstream architecture, in the form of services to share the resources related tothe processing of remote sensing products. Design the interface about Remote sensing production gridnodes and earth observation grid base on the production and real-time monitoring.2. Design the organization form of remote sensing production resources and access to EarthObservation Grid by interface server, and ultimately share the3-5level remote sensing production andstatus through the cluster parallel mode. Cluster parallel computing systems uses the strategy of reasonablycutting large remote sensing data to the small data, and then paralleling producing, which is Composed ofdata pre-processing servers, Order organization and management server, task scheduling server and parallelcomputing nodes.3Involved the realizing of Remote Sensing production grid nodes service prototype system throughgrid node architecture model and he organization form of remote sensing production resources. Theprototype system has been deployed on the servers of IRSA and accessed to the other grid node prototypesystem of Zhejiang University and Resources Satellite Center for trial production.