Study On Quick Retrieval Technology Of Multisource Remote Sensing Resources Based On Grid

With the rapid development of space technology in China, the research and development of remote sensing satellite has mature. In'12.5 plan'programming of our country, we should give great attention to the applications of remote-sensing data, and hope remote-sensing data can better utilize to the many fields. For example our national economy, evaluating natural disaster, scientific research and so on. Now we can receive a vast amount of remote-sensing data which usually have different types, different resolutions, different sensors. However these remote sensing data is often located in different places or belonged to different departments. At present there isn't an effective management system. And, as remote sensing data acquired by satellite is geographic information data, containing the geographic information, retrieval of the remote sensing data belongs to the spatial retrieval based on polygon area matching. And the calculation is usually quite large. In addition the databases contain a large amount of remote-sensing data. Because of that, it will take great difficulties for the remote-sensing data retrieval and retrieval rate is extremely low. So it is necessary to design a quick service system to these remote-sensing data.In order to break down barriers between the areas or departments and let the military satellite resources and the civilian satellite resources can constitute a virtual overall. In this thesis, we use Distributed Database System (DDBS), which connect different databases that are in distributed geographic locations, to manage these remote-sensing data that distributed in different places. And it realizes remote-sensing data sharing and user transparency.So that users can get the remote-sensing data quickly, in this thesis we realize a quick retrieval service system. In the quick retrieval service system, we take different ways to various problems. Firstly, about the complexity of the spatial retrieval problem, we propose a spatial retrieval method based on two levels. The mainly idea of the method is that it adds attributes retrieval on the base of spatial retrieval. So it will decrease the data that needs spatial retrieval and the retrieval speed will get a boost. Then, in order to satisfy the mass computing resources that are needed for spatial retrieval of the mass remote sensing data, we introduce the gird computing technology. It can share the distribution computing resources and offer tremendous process ability to system. In addition, the tasks scheduling algorithm is key technology in the grid system. In the thesis, we adopt Ant Colony Optimization (ACO) algorithm and we propose an improved Ant Colony Optimization algorithm. The improved algorithm can be better performance in load balancing of grid system. Lastly, so that the retrieval result can be better show to user, we propose a sorting method based on prestige degree.In this thesis, we develop a rapid retrieval system based on grid computing for multi source remote-sensing data and have an in-depth theoretical research. The system gives an effective management to the distribution and heterogeneous remote-sensing data. It also makes different area applicants can quickly get accurate remote-sensing data. The research of this thesis has a significant role in comprehensive application of remote-sensing resource.