Spatial Database Updating Using Map Conflation Techniques -- Road Map Example

• Main Authors: Lee Jieh-Ren, 李介人
• Other Authors: Sun Bing Leung
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/92673036091537417128

碩士 === 中正理工學院 === 軍事工程研究所 === 88 === The development of earth resources must have correct and timely information for planning. Due to the advent of high resolution satellite imagery, the changes of earth surface by natural and artificial power can be detected almost simultaneously. By the tools of digital photogrammetry workstation, digital mapping system, and geographic information system, the extraction of ground features and update of the geographic data base become feasible and also necessary to ensure correctness and timely information. Depending on the main components of map conflation, nearest neighborhood searching, Delaunay triangulation, rubber sheeting, provide the function to ensure the original valuable data can be kept, and the out-dated data can be updated timelessly. In this article, the history and the processing procedures are briefly discussed. The map conflation techniques is tested by using the data from two sources, the TM2°network digital data produced in 1991 and UTM network digital data in 1995. During the research, we found that if the nearest neighborhood searching distances were set to 40 meters, the searching would yield the maximum number of matching points. We also developed a node-based cutting technique to increase the successful transfer to about 90 percents, from TM2°network attributes to UTM network. At the last an iterative process flowchart was built for processing the other maps of the same series, and some conclusions were also drawn from the experiment.