The Study Of Multiresolution Vector Geographic Data Organization For Storage Optimization

As an important spatial data, vector geographic data is not only widely used, but also faces a series of challenges. On the one hand, with the enthusiasm of public participation, the request of vector geographic data service quality is higher; on the other hand, because of the restriction of hardware processing capability and the limited network bandwidth, there is a huge gap between vector geographic data service and real-time user experience.Progressive transmission and vector tiles are two mainstream technologies of improving the response speed of vector geographic data service. Progressive transmission using incremental transmission technology to decrease network data transmission, suitable for the client with independent drawing and caching capability; vector tiles converts the vector data to raster images, and it is suitable for all kinds of clients, while it has shortcomings, such as large transmission and storage and Symbol curing, and real-time tiles can partly solve the above problems. Based on vector geographic data multiresolution organization on the server side, progressive transmission and real-time tiles for vector data is provided strong support. Pyramid method is a common way of multi-resolution vector geographic data organization, but it can only adopt the fixed mode and cannot proceed fine-grained adjustment according to the spatial distribution characteristics of vector geographic data, thus storage efficiency is affected. In addition, the widely use of floating-point data storage structure also leads to large storage and affect the efficiency of the I/O.In this paper, starting from the goal of optimization storage to improve the efficiency of I/O and based on the thought of integer transform compression, we put forward the multilevel difference compression and real-time decompression method of vector geographic data; we put forward a kind of multi-resolution vector geographic data organization model based on STR-M tree, this model proceeds adaptive partition data space according to the spatial distribution characteristics of vector geographic data, and then do fine-grained adjustment for each unit of multiresolution organization scheme. In this paper, the main research work includes the following several parts:(1) Multistage difference vector geographic data compression method is proposed. The compression method makes full use of the correlation of orderly point set to do multistage difference transformation, combined with the variable long integer storage, realizing high compression of geometry information and express decompression, and satisfying the requirement of real-time decompression. This method is lossless compression effect at the specified resolutions, especially suitable for the compression of multi-resolution vector geographic data storage.(2) In view of the inherent disadvantages of the traditional pyramid scheme, firstly we establish corresponding STR-M tree partitioning scheme according to the spatial distribution characteristics of vector geographic data, and then We independently determine the multiresolution organization scheme and compressed parameters of STR tree leaf node (the division units) to optimize the storage efficiency, finally we based on STR-M tree to set up multi-resolution vector geographic data organization model.(3) We study the multi-resolution vector geographic data index structure and construction method based on STR-M tree, setting up a indexing algorithm which is based on the client window features to proceed the accurate positioning, high efficiency and rapid transformation for multi-resolution vector elements, and studying the adding and removing mechanism of STR-M tree.Based on related theory and method proposed in this paper, the paper establishes the prototype experiments and tests. The experimental results show that the proposed compression method and the multi-resolution vector geographic data organization model is superior to the traditional methods, obviously in the aspect of improving the I/O efficiency of vector geographic data, and it provides strong support for the progressive transmission and real-time tiles of vector geographic data.