Research On Multi-scale Concatenated Update Methods For Map Data

In the past two decades, the map digitization technology has achieved great development, and has been widely applied in the most cities of China. With increasing demand to this technology, the map data sharing and keeping them updated in real time had gradually become pressing problems. To maintain the data up-to-date the map database should be synchronized with urban and rural development. As is well known, in China today, the construction of most cities is changing with each passing day, and then updating map data became a fairly long and complex process. For a long time the map data updates were accomplished by a great deal of artificial participation with a low level of automation. The fully automated update of geographic data is still a very difficult task.General map data update not only means maintaining the specific scale map in keeping with the real world, but also includes fulfilling the concatenated updating among several different scale maps to ensure data accuracy and consistency. This paper focuses on the key technologies to utilize large scale map data for synchronously updating space elements in small scale maps. There are three key issues in the process of synchronization:automatic detection of graphic changes of the map, the choice of remove or replacement for these changes, concatenated updating for small-scale map data based on the updated results of large scale map data.To address the three main issues, this paper carried out a series of specific research.1) This paper clearly pointed out the objective, significance, mechanism of map data updating and a series of processing steps; reviewed developments of some key technologies involved in computer-assisted cartography, such as map data updating, spatial object matching, change detection; based on the analysis and comparison, this paper summarized the main characteristics of several map updating strategies at present, and proposed an incrementally map-updating technology, which is efficient to updating the multi-scale geographic information database.2) In order to accomplish database updates and historical data storage, this paper improved the basic state of amendment model, divided the databases into three classes according to the time phases, i.e., current database, temporary database and historical database. The relationship between the three databases is described, and the logical and physical designs are conducted subsequently.3) The fundamental prerequisite of change detection is identifying multi-scale counterpart elements. With consideration of the characteristics of multi-scale spatial object matching, this paper introduced multi-scale line-feature object matching utilizing Hausdorff distance and multi-scale area-feature object matching with consideration of influences by cartographic generalization.4) This paper summarized the characteristics of all kinds of map spatial data change, presented the formalized representation methods, and analyzed the main factors influencing the change-detecting caused by cartographic generalization. The processing methods on solving cartographic changes are pointed out in the operations of selection, simplification, aggregation, segmentation and collapsing.5) On the basis of analysis on change-detecting operation, this paper proposed overall plan of finding each kind of element change under the situations of identical scale and different map scales. This method could help to efficiently identify the element set to be generalized and the element set to be deleted.6) On the summarization of the current extracting methods of roads, this paper introduced the method to extract local road zone based on influence area, which fully utilized the existing map data and could remarkably reduce the computation load of extracting roads in the change-intensive areas. In order to solve the problems of spatial conflict and visual conflict, a new hybrid data structure of raster-vector is applied into the building simplification and mergence, which improved the traditional raster-vector data structure to meet the needs of detecting graphic conflict and merging buildings.7) The two different sets of experiments on the updating operation under two conditions, identical scale and different scales, are conducted to evaluate the performance of the increment map-updating technology. The experimental results showed that the proposed technical scheme is practical and effective, and a series of key methods involved in this scheme were proved to be accurate and stable, including multi-scale spatial object matching, change detection impervious to cartographic generalization, local road extraction method based on influence area, and the buildings-merging method based on hybrid raster-vector data structure.