Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Research and Application on Technology of Integration and Processing of Multi-Source and Multi-Scale Spatial Data (IPMMSD) XIAO Jihua, SUN Qun, LIU Haiyan Institute of Surveying and Mapping, Information Engineering University, 66 Longhai Road, Zhengzhou, China, 450052 [email protected] Abstract: this paper’s goal is to make fuse, manipulation, management, maintenance and service of spatial data to an organic integrity by IPMMSD, to digitalize model of geo-spatial data, also to bring an software platform for abundant spatial data’s management, to provide users an intelligent platform for getting spatial information rapidly, and also to make an basic software for spatial dynamic and quantitative analysis. On the support of spatial database management system to realize integration of multi-scale, multi- resolution, multi-type, multi-time digital map, image and property data, and manage these data by concentrative or distributive way, to make abundant isolated information into share data resource. Keywords: Multi-source Spatial Data, Multi-Scale, Spatial Data Model, Spatial Data Engine, Spatial Index, Map Symbolization, Map Database, Integrity of Spatial Data. 1 Introduction In 1992, GoodChild put forward the concept of geographic information science, it regarded that geographic information science should study the information flow of geographic information system, its main goal is to solve such basic problems as manipulation, storage, extract, management, analysis of geographic information. But the development of GIS (science) is formed from cartography. Because of the rapid development and comprehensive application of computer technology, space technology and communication technology, it realize the ultimate revolution from traditional map-making manually to digital map-making and survey. At the same time, the technology of earth information collection moved from geodetic survey to automatic remote sensing technology, and the manipulation of geographic information is gradually extended from graphic show to storage, analysis, simulation, prediction and share of geographic information. So the main research direction is changed from traditional paper map-making to IPMMSD. 2 The Urgent Need for Multi-source Multi-scale Spatial Data Integration Processing In recent years a growing number of departments, units and personal needs at the macro, meso and micro scale, using various multi-scale digital geographic information on the Earth's surface all the natural and human phenomena at different scales of space morphological structure of the abstract expression, forming a series of topographic maps of different scales, to establish the basis for the corresponding geographic database to enhance the planning, management, monitoring, efficiency and level of decision-making (Jun Chen, 2000; HeHai Wu, 2000). In order to fully utilize the rich data resources, to different levels of management, analysis, planning and decision-making content and gradually increasing detail and accuracy of spatial data to an urgent need for an effective approach to multi-source multi-scale spatial data processing and management, through the model, geometric structure, semantic interpretation of the level of expression and encoding appropriate conversion process, different scales from different sources to integrate spatial data were analyzed and compared. Spatial Supported by National Nature Science Fund(41071297) 28 database management system in support of multi-scale, multi-resolution, multi-type, multi-temporal vector maps, images and attribute data integration, to the past, a large number of scattered data into easy to share vast amounts of data resources. 3 The Main Content of Multi-source Multi-scale Spatial Data Integration Processing 3.1 Multi-source multi-scale spatial data integration with data model Taking into account the multi-source multi-scale spatial data integration processing and map the dual task of visual display, multi-source multi-scale spatial data integration processing system also uses two hierarchical models. Is like a house of the two windows. But not the same hierarchical organization by mapping to control the order of map symbol gland production to meet the traditional paper maps the requirements of the visual habits, and geographical stratification is conducting spatial data integration is an index of processing mechanism, it does not affect the relationship between elements of the gland and the imprint. That is the same two layer model is spatial data indexes of two different ways, depending on spatial data processing and cartography of the different requirements of the conversion through a simple interface to achieve "Viewport exchange (data change)." Figure 1 shows a view that contains two different conceptual models of spatial data retrieval. Multi-source multi-scale spatial data integration, integration of processing and cartography, GIS and Digital Cartography and the advantages of two types of spatial data model and its data processing method is proposed in this paper on the multi-source multi-scale spatial data integration and processing of business processes spatial data model, which take into account the spatial data processing and analysis and visualization of map drawing. Layer1 View For Geographic Information Layer… Layer1 LayerN GeoDataSet1 GeoDataSet….. ( View for Map Graphics of Geographic Information LayerN GeoDataSet N Formalization Geographic Information GIS Model Map Database Layer… Point Line Polygon ) Classification for Geographic Information GIS Data Controller Annotation Complex Geometry DEM Image ( Map Map Model Map Symbolization Controller ) Cartographic expert’s knowledge Formalization MapDataSet 1 Layer 1 Layer… MapDataSet….. Layer N MapDataSetN Layer 1 Layer… Layer N Figure 1 Multi-source multi-scale spatial data processing and cartography integrated conceptual model 3.2 Multi-source multi-scale spatial database design In the traditional system of geoscience applications software commonly used documents and manage 29 map data stored. As compared with document management, spatial database technology in the mass data management capabilities, graphics and attribute data integration, storage, multi-user concurrent access, and improve access control and data security mechanisms and other aspects of the obvious technical advantages. Spatial database technology is gradually replacing traditional file management, a growing number of GIS applications and medium-sized data storage solution space [2]. Follow Oracle's data management system, database structure design as follows: Dictionary tables and data tables will be divided in different table spaces; different types (including different scales and sources) of data is also on a different table space; such as table space TS_1000W 1:1000 million world map data storage, TS_300W Storage 1 : 300 million Chinese geographical map data, TS_100WK 1:100 million aviation map data storage, TS_25W 1:25 million collaboration diagram store data, TS_230WH chart data storage 1:230 million, TS_GPS stores nationwide road traffic data, GPS, you can set the table storage space for the automatic growth; and according to certain criteria for the classification of each scale data will be divided into several categories, such as industry and agriculture, and social and cultural facilities, residential areas and ancillary facilities, land transport, for each class is divided into a point, line, surface, annotation, composite, metadata and other data tables. Table 1 The same scale map composed of data tables Layer Name The composition table T_25W_B_POINT T_25W_B_LINE T_25W_B_POLYGON Industrial and agricultural and socio-cultural facilities T_25W_B_COMPLEX T_25W_C_POINT T_25W_C_LINE T_25W_C_POLYGON Residential and ancillary facilities T_25W_C_COMPLEX T_25W_D_POINT T_25W_D_LINE T_25W_D_POLYGON Land Transport T_25W_D_COMPLEX …… ………… 、 、 、 、 、 、 、 、 、 、 T_25W_ANNOTATION T_25W_ANNO_COMPLEX Notes 3.3 Multi-source multi-scale spatial data engine design MapVisualization Edit Query Analysis Other Application…… Application National Map Data SDE SDE Data Exchange 、 ArcGIS E00 shape data OracleAccess SQL Server Access MapInfo MIF data …… Other spatial data……… Oracle SQLServer Access DB2…… Data Figure 2 Spatial Data Engine data source Spatial database engine is not the spatial data, but a bridge, would use and storage of spatial data linked to the data depends on the relational database for storage and management, but application of spatial data through spatial database access Engine to achieve. Spatial database engine is to provide storage, 30 query, retrieve geospatial data, geographic data and the spatial relationship of space and spatial analysis operations procedures set of features. For example, the main vector spatial data engine component of Figure 3 can be expressed: Figure 3 Vector spatial data engine class diagram CMM_SpatialEngine object is the spatial data engine for all objects in the root, it can traverse through the multi-scale map of access to all data objects in the database, including all map types, map, layers, elements of sets, geographical factors and their geometry, attributes and topology information. A CMM_MapType object maintained in memory of the type of point, line and plane map symbol library, color library, coding system, coordinate system, projection, coordinate units, and names of geographical elements symbolic note, the property that the sign of the note control information, access to data structures and their interfaces. CMM_Map class object represents a map database, which contains the description of the map-related information, and it contains all the geographic layers objects and publishing layer objects, and provides information and a description of the map contains all the layers object traversal methods. CMM_Layer class object represents a map object belongs to a vector geographic layer, which contains the description of the layer, the layer structure information of the field attribute data, and the layer contains the geographic features of the object composition kinds of data sets. Also provides access to these data. CMM_PubLayer object of the class, mainly from the perspective of the traditional paper maps to consider the publication of geographic information (map symbol) hierarchical, it is mainly to control the map display to map symbol relationship between the gland in order to enhance the map display and clarity. CMM_FeatureSet class object represents a layer in a geographic element (CMM_Feature object) of the collection. It is responsible for maintaining a pointer to the object list of geographical factors. CMM_UniFeatureSet object of the class that has the same geometry type of geographic features a collection of objects. It is responsible for maintenance of a single type of pointer to the object list of geographical factors, such as point, line, surface, composite elements, annotation objects such as a type of geographic features a collection of objects. CMM_Feature class object represents a real-world geographic features. Element object can be a geographic point, line, surface, composite elements, such as five different note types, geographic elements with the attribute data, geometric data, topology data, information and notes symbolic information. CMM_Attributes object encapsulates a geographic element object attribute data, attribute data and provides the read 31 operation. CMM_FieldsInfo object encapsulates the attributes of a field layer structure of geographical information and its access method. CMM_GeoPoint, CMM_GeoLine, CMM_GeoPolygon, CMM_GeoComplex object encapsulates the points, respectively, line, surface, composite elements of geometry and topology information. 3.4 Spatial data access and index 3.4.1 Control of memory data All the data of Single map is read into memory for once and for all. As for the pieces of map data, the situation becomes complicated, the map display window changes to be dynamically updated map data in computer memory. To map the roaming operation, for example, as the viewpoint moves, the contents map window with the map data on the critical updates, it is clear view of the constantly moving, changing the map display range from the need to continue to map database loaded with new data, when roaming constantly increasing the number of memory data is updated constantly changing, if only just keep the new data into the memory consumption will gradually increase, and will soon be fit new data, so changes in the map window, in addition to reload from the database, the new map data, but also need to release the previous window is displayed, but was removed from the current window contents, so there is increasing have less to maintain the memory map data relative balance. Data management, multi-map maps the basic idea is simply the map window with the dynamics of vision and efficiently implement map data between disk and memory scheduling. 3.4.2 The establishment of grid index Using the thin dotted line to indicate on each map sheet have established a 2 × 2 grid index. Clearly, then when the window moves from the viewport to viewport 1 of 2 areas, would not have transferred from the database of the three maps Figure 1,2,4 complete data, and only a quarter of each image One can, obviously the data scheduling efficiency has been greatly improved. By adjusting the density of grid index can be a relatively satisfactory data scheduling efficiency. Established for each map sheet grid index, despite the increase in the roaming map sheet map of the efficiency of the junction, but within a map when roaming from one index to another index of the grid need to access the database transferred to the grid new data. Grid index should be established to consider the following two key questions: 1. Grid density level In the real space grid index is created, is to divide a layer for each grid. Experimental results show that the index in order to get efficient results should follow the following principles: the different layers were established grid index; grid data by the general principle is that a number of large dense layers, layer a small amount of data sparse number; but because the division is not the more dense network the better, generally 10 × 10 (experience) is appropriate; very small amount of data layers do not need to create a grid index. viewport2 1 2 viewport2 3 4 1 a viewport1 b 4 viewport1 Figure 4 Grid index diagram 3 2 5 c 6 Figure 5 To establish elements of updates across the data grid 2. Updates across multiple data elements in the treatment of grid Shown in Figure 5, black thick lines that map the scope of one of these factors created a layer of 2 × 3 of the grid index, were numbered 1,2,3,4,5,6, which line 1,2,3,6 a span of four grid, line b only 4 in a grid, 32 line c across 2,3,5,6 four grid, depending on the port 1 (Grid 1 and beyond 2) that the previous map window, and after a roaming to the viewport that the window range of 2 (complete in grid 2). The following detailed analysis of such a window to change the memory caused by data changes. When the window in viewport 1 state, a data grid is loaded, and record the address of a line, while 2 of the data grid is loaded and the address of record c, where the need to pay attention, even though a line across the two grid 1 and 2 but in order to avoid duplication of data in memory, the line a record by a grid 1 grid 2 will no longer be recorded. Therefore, when the window is roaming to the location of the viewport when 2 (completely fell to the grid 2), memory data updating mechanism will be judged not on the map grid 1 has been displayed in the window to make its records all elements of all the clean out in memory, when the window's position in the dashed box grid 1 and grid 2 data is loaded, meaning that 2 of the data grid has been set to true load flags, so When the window to viewport 2 roaming location, the grid elements of 1 is to be cleared away, and 2 of the data grid is not updated, because its data is already loaded flag is set to true , so this result is already roaming should also be in-memory and displayed elements of a line memory is cleared out, so will not see, obviously, this is a result should not be there. To avoid this error delete the elements of the results, it must clear a space in the network 1, one by one to determine when the memory data grid elements of the external 1 and grid 2 rectangles are overlapping occurred, if there is overlap, it can not deleted, but to address the elements of the transfer to the grid 2, so you can avoid mistakenly deleted elements a. 3.5 Multi-scale spatial data and interactive control of symbolic The purpose of the map is a visual map of the visual transmission to meet the needs of particular users for most of the public key used to express a known phenomenon, and its ability to interact with low or no interaction. For the application of professional personnel, such as scientists, engineers often need a strong ability to interact with map visualization systems, because the experts needed to explore the cognitive analysis of the geographical space of things the new law, which is more meaningful than the visual transmission of visual process. Digital map symbols from the map of knowledge in the expert's experience and knowledge that is loose and a lack of logic between the close contact. Therefore, how to extract useful knowledge from these information for the map visualization is particularly important. Therefore, including user interoperability knowledge base, including more detailed process can be described under Figure 6. Map Expert Programmer express Developing Phase Knowledge Need Computer Rule study Restrict User Using Phase Formalization Knowledge base User customization Figure 6 Expert knowledge and user self-learning Figure 7 Bridge attribute Annotation In digital cartography system development phase, requires the user, map experts and system developers to consult with the needs of the user, the characteristics of the computer to achieve a certain constraint conditions in the establishment of an expert knowledge base. After completion of the system, the user can modify the Knowledge Base by editing some of their own intentions and ideas established in accordance with the constraints of knowledge (systems provide a set of constraints input to ensure the correct interface) into the knowledge in knowledge, in order to complete specific application. For 33 example, a bridge between different levels of road signs that they are different widths, and can be abstracted into a formal knowledge of the following: Table 2 Highway bridge symbolic formal representation of expert knowledge Road Bridge 140523 PO [type]="state road" 140523 Road Bridge 140523 PO [type]="high way" 140523A Road Bridge 140523 PO [type]="province road" 140523B Table 3 Notes to control the map's expert knowledge of expression in the Knowledge Base Railway 140522 PO Middle Normal 7 Black {{[Len]-[Wid]}/[Load]}@[Type] Bridge Bold 0 The user interface can be achieved through the appropriate note on the custom properties of the formal entry symbolic knowledge. 3.6 Multi-source multi-scale spatial data integration approach Multi-source, multi-scale spatial information collection to the space involved in a wide variety of data sources, however, these spatial data access ways and at different times, making them the scale, mathematical foundations, semantic definition, geometric representation, data structures, etc. the existence of many inconsistencies, to the application of multi-source spatial data bring great difficulties. In practice, you can handle a variety of methods of spatial data summarized as the following aspects: 3.6.1 Pretreatment of multi-source spatial data Pretreatment of multi-source data including vector map data format conversion, map projection and coordinate system conversion transformation, and establishment of topological relations. 3.6.2 Multi-source multi-scale spatial data matching techniques Practice shows that the integration of geographic information from multiple sources and not just a simple data format conversion or coordinates, Projection Transformation, etc. will be able to meet the requirements, which need further spatial data matching technology to solve [7]. Spatial data matching and the matching of different entities, including entities matching the same name. Different entities of the match, mainly for the GIS analysis applications such as databases 1:100 million residents of the road data and ground data, can be obtained by matching the number of a road connecting more than 50 million residential population to determine the The importance of the road. Named entities is through the analysis of spatial entities, matching the differences and similarities of different sources to identify the expression of the real world map surface features or surface features the same set (ie same name, entity) process [8]. 3.6.3 Integration of different data encoding techniques A vector for each data source has its own coding system, if you want to use the data to digital map into their products, must be coded system conversion. This requires in-depth analysis of different data encoding system of standards, targets the same name in different industry standards and their attributes in the code corresponding to the target type and corresponding attribute the contents of the establishment of semantic matching between the two models. Table 4 Million a 1:100 data standard data standards 1 and 2 encoding the corresponding table Data Standard 1 Data Standard 2 Content Code Layer Geometry Layer Content Code Geometry River(mongline) 21011 H Line F River 160201 Line River (crewel) 21012 H Sign F River 160201 Surface field Point Point River (crewel) 21021 H Line F Bank Line 160101 Line Seasonal 21022 H Line F Seasonal River 160202 Line 34 River(mongline) Seasonal River(crewel) Lake ……………. 4 21022 H 23011 …… H …… Sign Point Line …… F Seasonal River 160202 F …… Bank Line …… 160101 …… Surface field Points Line …… Conclusion and Outlook Integration processing is the integration of spatial Shuju the initial stage, is still just stop at the relatively mechanical, simple centralized management and display stage, there are not enough ways and means to let the computer replace the human brain to analyze the different sources and scales the difference between data and links to discover the useful knowledge contained in them. The automation of cartographic generalization is difficult at this stage, you can consider multi-scale spatial data processing technology has been updated from a study of large-scale map data to update obsolete small scale data (such as 1:25 million and 1:50 million map databases have been established, as long as 1:25 million to update the library can go through the multi-scale processing technology automatically updated 1:50 million map), a vast country like China that the big countries build the basis of geographic information can save a lot of human and financial resources. References [1]. Xiao JiHua. Multi-source multi-scale processing of spatial data integration and application of [D]. PhD thesis, Information Engineering University, 2008(in Chinese). [2]. Beijing GIS Technologies Co., Ltd. Super. SuperMap SDX + Technical White Paper .2003.6. (in Chinese) [3]. Xiao JiHua, Liu Yan, Luan Xiaoyan. Geographic Information generation and integration of the conceptual model of cartography, mapping projects, Volume 16, December 2007 No. 6 :5-8. (in Chinese) [4]. Xiao JiHua, Sun Qun, Liu Yan. Cartographic expert knowledge of geographic information, symbolic control of the process. Surveying and Mapping Technology. Volume 24, June 2007 3, 189-192. (in Chinese) [5]. Xiao JiHua, Liu Yan, Luan Xiaoyan, Zhou Zhao. Open data management integration vector map design. Surveying and Mapping Technology, Volume 23, June 2006 3:198-201. (in Chinese) [6]. Xiao JiHua, Liu Yan. Multi-source multi-scale map data organization and management, Surveying and Mapping Technology, Volume 26, February 2009 No. 1:24-28. (in Chinese) [7]. Xin Wang. Multi-source spatial data matching method [D]. PLA Information Engineering University, Master Thesis, 2008. (in Chinese) [8]. Filin S. and Doytsher Y.. A Linear Mapping Approach to Map Conflation:Matching of Polylines[J], Surveying and Land Information Systems, 1999,59(2):107-114. 35