Design And Implement Of Massive Remote Sensing Image Management System

With the rapid development of remote sensing technology, the applications of remote sensing data have been in all walks of life, such as weather area, military field, forestry area, urban planning area, etc. Along with these phenomena, the amount of remote sensing data is increasing and the rapid growth is of magnitude from the GB to the TB level.In the face of massive remote sensing data, how to effectively manage and how to find and obtain useful data, this is a big problem. Each institution encounters enormous difficulties and challenges and a practical large-scale remote sensing image management system is urgently needed. With the rapid development of information technology, the state-of-the-art technologies are also rising on the large-scale data processing which have been providing people with new ideas and solutions. In particularly, Hadoop distributed file systems(HDFS) and MapReduce programming framework proposed in recent years have a good capacity for certain large-scale data.Based on the massive remote sensing data management challenges, an approach to manage large-scale remote sensing data explored and researched in this paper and then a massive remote sensing data management system was designed and implemented. To achieve an effective management of massive remote sensing image, a large scale image storage and a fast way of retrieval, access and browsing must be needed. The system is the one combined a distributed file system and relational database management system in the paper. The original data and its metadata are stored separately. The original large-scale data is stored in distributed cluster and the metadata in a relational database. For the remote sensing data, the metadata management will turn to be a very important part of the system. Therefore, the metadata management is treated as a separate subsystem. For a large-scale remote image, the job of metadata extraction is also a challenge. So, the use of MapReduce on Hadoop platform to efficiently extract the metadata from massive images is focused on in the paper, which can improve the efficiency on metadata extraction of massive images. Finally, a prototype system named MRIM_MS of massive remote sensing image management was designed and implemented, which has features such as multi-resolution, fast retrieval and 3D visualization browsing.