Matching planar maps

The subject of this paper are algorithms for measuring the similarity of patterns of line segments in the plane, a standard problem in, e.g., computer vision, geographic information systems, etc. More precisely, we define feasible distance measures that reflect how close a given pattern H is to some part of a larger pattern G. These distance measures are generalizations of the well-known Fréchet distance for curves. We first give an efficient algorithm for the case that H is a polygonal curve and G is a geometric graph. Then, slightly relaxing the definition of distance measure, we give an algorithm for the general case where both, H and G, are geometric graphs.  2003 Elsevier Inc. All rights reserved.

• Type: Conference
• Year: 2003
• Where: SODA
• Authors: Helmut Alt, Alon Efrat, Günter Rote, Carola Wenk
• Keywords: Algorithms, Distance Measure, Feasible Distance Measures, Geometric Graphs, SODA 2003