| The explosive growth of remote sensing image data poses a huge challenge to the storage and retrieval of data.In this paper,based on the research and analysis of relevant architecture and technology,a new solution is proposed for the management requirements of massive remote sensing image data.Finally,a remote sensing image data management system is designed and created.The main research and conclusions of this paper are as follows:(1)Compare and analyze the current mainstream distributed file system,and select HDFS as the storage system of the remote sensing image data entity.In view of the shortcomings in storing massive remote sensing images,redundant backup and cooperative work are used to solve the single point fault of NameNode to ensure the fault tolerance of the system.The monitoring system is introduced to monitor the performance of the node and alarm the system to improve the system.Stability.(2)Contrasting and analyzing the current mainstream NoSQL database,selecting distributed search engine to organize the storage of remote sensing image metadata.Proposed based on Elasticsearch,The MapReduce model is used to automatically extract multi-source heterogeneous remote sensing image metadata metadata,and the large-scale metadata is indexed by pg-es-fdw mechanism to realize efficient retrieval of data.The experimental results show that the retrieval speed of this method is significantly better than the traditional relational database in terms of retrieval speed.(3)Based on the image storage and organization management methods proposed in this paper,analyze the system requirements,determine the system architecture,and design and develop the remotesensing image management object in the topic of “Geographic Resources and Ecological Professional Knowledge Service” of China Engineering Science and Technology Knowledge Center.system.The system commissioned a third party to test its functionality,performance efficiency,reliability and information security,and all the tests passed. |
PDF Full Text Request