A Parallel Factor-Based Method of Arterial Two-Lane Roads Recognition

Urban arterial roads recognition and extraction is an essential step in networks generalization, and the handling of two-lane roads is a problem in large scale map generalization. Aiming at urban arterial two-lane roads recognition, in this paper, constraints about candidate pairs of two-lane roads are built based on Gestalt principles, and a parallel factor-based method of arterial two-lane roads recognition is proposed. Firstly, a topology between the nodes and road edges is built. Secondly, the matching method of Hausdorff(HD) distance to recognize candidate pairs of two-lane road like roads matching is utilized. Thirdly, computing candidate pairs parallel factor, if the parallel factor satisfies the threshold condition, the pairs are sections of an arterial two-lane road. Finally, according to the spatial relationships, many recognized sections are connected into whole roads. Experiments show that this method can effectively extract arterial two-lane roads in the road network after setting the threshold value by selecting typical samples.