American cities, especially their central regions usually have a very regular street pattern: We are given a rectangular grid of streets, each street has to be labeled with a name running along its street, such that no two labels overlap. For this restricted but yet realistic case an efficient algorithmic solution for the generally hard labeling problem gets in reach. The main contribution of this paper is an algorithm that guarantees to solve every solvable instance. In our experiments the running time is polynomial without a single exception. On the other hand the problem was recently shown to be MV-hard. Finally, we present efficient algorithms for three special cases including the case of having no labels that are more than half the length of their street.
CITATION STYLE
Neyer, G., & Wagner, F. (2000). Labeling downtown. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 1767, 113–124. https://doi.org/10.1007/3-540-46521-9_10
Mendeley helps you to discover research relevant for your work.