Directional relations are fundamental to spatial data queries, analysis and reasoning. Consequently there has been a significant amount of effort to determine directional relations between two regions. However, many existing methods do not perform well when the regions are neighboring or intertwined. In this paper we introduce a new model for directional relations which is based on a splitting line separating the two regions in question. We identify essential quality criteria for directional relation models and translate them into measurable properties of a given splitting line. We present an efficient algorithm that computes an optimal splitting line for two regions and perform extensive experiments. Our results show that the splitting line model captures directional relations very well and that it clearly outperforms existing approaches on pairs of neighboring or intertwined regions.
|Title of host publication||Proceedings 19th ACM SIGSPATIAL International Symposium on Advances in Geographic Information Systems (ACM-GIS 2011, Chicago IL, USA, November 1-4, 2011)|
|Place of Publication||New York NY|
|Publisher||Association for Computing Machinery, Inc|
|Publication status||Published - 2011|