Remote Sensing Tile Organization Method And Workflow Calculation Model For Real-time Map Service

As a good way to display data sets,map services play an important role in the sharing of geographic information.However,the existing map services focus on content sharing and visualization,and there are few researches for real-time analysis and processing of images in the map service,and it is difficult to meet the current diversified real-time remote sensing analysis,result sharing and other integrated scenarios.The more value of remote sensing data is to make full use of its spectral information to analyze the characteristics of the geographical environment,so as to better serve the natural environment and human society.Therefore,how to quickly obtain image data from the cloud and carry out efficient information processing and value mining has become the focus and difficulty in the current remote sensing information field.This research uses map services to quickly acquire cloud image data and perform real-time analysis and processing,to achieve direct visual conversion from raw data to data products,and overcome the complex and lagging problems of remote sensing applications in processing,visualization,and sharing processes.To speed up the extraction,sharing and rapid application of remote sensing information.The specific research content is as follows:(1)An image data organization model(Zorder-Cloud Optimized Tile,Z-COT)combining Cloud Optimized Geo TIFF(Cloud Optimized Geo TIFF,COG)and Z-curve is proposed to continuously store adjacent grid tiles in space.And record the offset and size through the index file,which solves the problem of poor data transmission efficiency based on cloud image tiles in the area of interest.At the same time,a strategy that adapts to pyramid data reading is adopted for Z-COT to meet the requirements of different levels.For the data needs of map grid tiles,the map service's access to image data is improved from two aspects.(2)According to the request rules of map services,a distributed collaborative prefetching strategy for grid tiles is designed,and the message middleware is used to asynchronously load the neighboring tiles requested by the current map service,so as to realize the cooling of the grid tile data.Thermal hybrid reading further optimizes the efficiency of obtaining image data from the cloud at the system level.Aiming at the problem of repeated execution of distributed environmental message tasks,the concept of cache barrier is proposed and this problem is effectively solved.(3)On the basis of the efficient loading of raster tile data,an expression-based raster tile processing model is proposed.By converting the expression into a computational workflow,it is possible to realize the correctness in the request of the map service.Real-time processing of raster tiles enables rapid analysis of massive remote sensing data stored in the cloud.For scenarios where full data is involved,use appropriate resampling data to simplify calculations to meet the real-time nature of map services.Experimental results show that the Z-COT organization model proposed in this paper can effectively improve the efficiency of obtaining image data from the cloud,and as the amount of data increases,the efficiency of obtaining data is more obvious.In terms of the efficiency of full data reading,it is 5 higher than COG Times.Using three different complexity models of NDVI,ground object classification,and vegetation coverage,real-time calculation and analysis of Landsat 8 images are carried out in the map service.It is verified that the workflow calculation model can effectively analyze the raster tiles and can be extended in a distributed manner.It can provide stable map service capabilities in high concurrency scenarios,adapt to calculations at various levels of scale,and provide a new idea for the development of map services in the future.