Architecture Researching On Remote Sensing Data Distribution Based On Spherical Partition

As the technology of remote sensing progresses, its data source becomes global, that faces huge global spatial data organizing and remote sensing information applying problem. Global climate research, monitoring flora and fauna study also exist similar problems. Traditional projection plane uses grid, which is suitable for small range statistics, Grid units represented by the ground position will become heavy asymmetric spatial distribution when grid Expands the scope of even the world. Scale conversion and wide-area or global statistics will have a larger error. At present, the location-based services (LBS) has promoted geographic information systems into the stage of information society services, From the department or industry-oriented services into the public service, from the map information Publishing into integrating the location of space-based information resources related to a variety of locations. The traditional model of spatial representation, collection, management and publication, which is based on maps, can not afford those new demanding.Global partition model which is based on global, multi-resolution, multi-scale transformation, evenly distribution, provides a spatial Index, uncertain located, gapless, spatial hierarchical, open represent integration. However, remote sensing image data is faced with a huge, distribution difficulties, slow update problem, so how conditions at the existing network to enable fast access to our latest high-resolution remote sensing imagery has become the focus of widespread attention. Remote sensing data distribution architecture based on spherical partition model is necessary to support the multi-scale, arbitrary size of the rapid distribution of image data, when some natural disasters such as earthquakes, floods happen, the remote sensing image data to the rapid application can locate the position disaster happened, at the same time be able to accurately access to disaster areas around the environment for the rescue organization to be able to quickly deal with the contingency provided strong protection, which has broad application prospects.The paper is based on Spherical partition model, builds a remote sensing data distribution structure. The main work followings:First of all, introduce the current domestic and foreign three popular spherical partition models: latitude and longitude grid model, polyhedral grid model and Adaptive grid model. Analyses various types of model inadequacies. Because of regular polyhedron with the adaptive mesh subdivision of the triangular patches for the many, complex geometric shapes such as polygons, rather than the regular rectangular or square, even though the expression of Convenience, but is not conducive, it is proposed the use of sub-sites based on the map partition of the spherical model, in order to increase the basic sub-scale map based on the use of latitude and longitude intervals of the level of global subdivision, a total of 25 scale subdivision. And subdivision surfaces have been precision at all levels to do the analysis. At the same time, this paper selected the order of Hilbert Space-Filling Curve subdivision code to mark all patches, the patch code can be maintained near the level of spatial relationships and to meet the world's massive spatial data management and efficient subdivision surfaces on the demand index. And to propose a code generation method of patches, the Hilbert curve based on the level of replication feature to partition the process at the level of direct labeled patches generated at all levels. Summed up after the model of remote sensing data at the distribution of advantages.Secondly, the use of tiles quadtree pyramid set up images of spherical mesh model data organization, analysis of tile first four plane-tree at the application of remote sensing data, and then extended to the spherical model of the application partition, give specific formula for calculating the change of the above formula will be Quadtree tiles all the tiles on the Earth is mapped to build a complete data model of the spherical subdivision. Remote sensing because of the very large amount of historical data, all depend on Quadtree tiles image pyramid will set up the server computing power have brought substantial challenges, in order to improve the spherical partition the data storage, in this based on the proposed incremental remote sensing image storage technology, through the establishment of a variety of three-dimensional space-time model and analysis of the incremental method of storage, remote sensing data to ensure the information is not lost, the compression of remote sensing data storage space.Then introduced the current domestic distribution system of remote sensing image data of the system architecture, the key technical characteristics of architecture, at the same time introduce more popular abroad adaptive middleware communications environment, including the architecture of the framework, the level of the structure and function of each level . After the proposed subdivision based on the spherical surface of the remote sensing image data distribution system architecture, designed the overall architecture of the distribution system and introduction of each module has the functions and tasks required. After the network communications environment using adaptive implementation of remote sensing image distributed middleware communications system architecture design, the use of connectivity - the recipient of the model, using multi-threaded, reactor technologies, such as set up a C / S of the data communications model, a detailed introduce the characteristics of each module, and said that explained the advantages of this architecture.Finally, in order to be able to better use of network bandwidth, transmission of information more useful to the application of remote sensing data terminals, remote sensing data in this article include the use of geographical coordinates of the characteristics of a location-based remote sensing data distribution strategy, focus on the application of remote sensing data to terminal provide the geographical coordinates as the center, the radius of the specified distance of a region of remote sensing images of the Change, to calculate the corresponding scale of the need to update the number of patches and the number will be sent to the application of remote sensing data terminals, to reduce the server and data transmission between the terminal redundancy, and remote sensing data give the flow distribution.