POI Selection in Multi-scale Maps Based on Voronoi Diagram
ISBN 978-85-88783-11-9
Authors
1Ying, S.; 2Chen, G.; 3Cao, X.; 4Li, C.
1WUHAN UNIVERSITY Email: shy@whu.edu.cn
2SHENYANG GEOTECHNICAL INVESTIGATION & SURVEYING RESEARCH INS Email: gqc@whu.edu.cn
3NAVINFO CO., LTD Email: caoxiaohang@navinfo.com
4WUHAN UNIVERSITY Email: 464851637@qq.com
Abstract
At the beginning, raw POI data have been increasingly collected by hand or via volunteered geographic information, without scale factors with various types. These POI data accumulate and are stored in database with a very large volume, waiting for be selected and visualized in certain scale map. Although every POI with spatial location and certain semantic attributes has its own value to be existed, not all the POIs can be selected and visualized in multiple scale maps due to our visual cognition, limitation of the display screen or the size of paper map as well as the map load capacity and information entropy. Thus, it’s critical to adopt the classification and graduation of POI data to satisfy multi-scale map visualization. Most studies about POI focus on position accuracy and attribute comparison separately with the spatial cognition and statistics rather than the combination of spatial relationship and attribute to visualize POI in multiple scale maps. POI classification and coding are discussed firstly with prominence analysis to manifest the importance of POIs. Then the paper proposed an algorithm for POI selection from raw POI data to adapt multiple digital map visualization, considering POI distribution mode in areal regions which are enclosed by roads. Voronoi diagram of POIs within a block are built to help to select POIs from raw POI data to adapt the visualization of multiple scale maps. The experiments illustrate the feasibility of the method to effectively satisfy the quantity and quality requirements in POI selection for multiple scale maps.
Keywords
Point of Interest; Voronoi diagram; Multi-scale maps