Method Of Remote Sensing Big Data Storage Management And Research On Interaction Visualization

The rapid development of remote sensing technology,providing a convenient for us to get remote sensing data.From acquisition to application,remote sensing data has experienced a transition from data to information.However,with the rapid growth of remote sensing data,it has brought challenges to the efficient and orderly storage and application.Traditional distributed file storage method is a very effective file storage management method,but frequent file access processes are bound to limit the efficiency of system processing and response if under high concurrent requests.Network transmission,visualization and application of spatial data has been an inevitable trend,also,the storage of remote sensing data inevitably restricts the efficiency of remote sensing data visualization applications.Relying on distributed file storage does not take the two parts of data management and visualization applications into account.Therefore,how to combine the two part together and use related technologies to accomplish remote data storage management and interactive visualization is the focus of this study.Based on Remote Sensing Big Data Storage Management Method and Computer Technology,building Remote Sensing Big Data Real-time Rendering and Interactive Visualization System,this article mainly completed the following aspects of the work:Achieve efficient access to massive image tile data.Base on the method of distributed file management system,this system designs redis cache server and build redis cluster.In the file access process,by relieving stress on namenode nodes,it can improve the access efficiency of image tile data and reduce the pressure on the server,finally archiving load balancing of the system under high concurrent requests.Efficiently realizes the linkage index of image tile data and its metadata.This article builds the physical storage path of the image tile data according to the meta information such as the name of the original image,the level,and the split row number,and designs an image metadata index database and a plurality of heterogeneous metadata information databases to store massive image metadata.When image data is rendered in real time,the server retrieves the tile file storage nodes step by step according to the parameters requested by the client through calculation and analysis.According to the index information in the image metadata index database,we find the rendering metadata in the metadata database,and finally realizing the linkage indexing process of image tile data and image metadata.Realizes real-time rendering and interactive visualization.Based on the current tile and geo-rendered tile data,by splicing,cropping and adjusting resolution of tile data,this system provides users with real-time rendering of multi-scale image data and interactive visualization service process,including image data retrieval,image data download,user file upload,user file management,metadata interaction service interface and other main functions.This article analyzes the system data from the file access rate,load conditions,and interface performance.Compared with the original distributed file storage method,the storage efficiency of the file is improved;based on the number of server connections and performance load,optimizing the system load balancing policy,greatly reducing the pressure on the entire server.