2009/09/11 by Michael A. Bekos, M. A. Bekos, Michael Kaufmann +5 · 1 citation
Computer Science · Engineering · #Computational Geometry and Mesh Generation #Digital Image Processing Techniques #Advanced Numerical Analysis Techniques
paper · doi:10.1093/comjnl/bxp087
Boundary labeling is a relatively new labeling method. It can be useful in automating the production of technical drawings and medical maps, where it is common to explain certain parts of the drawing with text labels, arranged on its boundary so that other parts of the drawing are not obscured. In boundary labeling, we are given a rectangle R which encloses a set of n sites. Each site si is associated with an axis-parallel rectangular label li. The labels must be placed in distinct positions on the boundary of R and to be connected to their corresponding sites with polygonal lines, called leaders, so that the labels are pairwise disjoint and the leaders do not intersect each other. In this paper, we study a version of the boundary labeling problem where the sites can “float ” within a polygonal region. We present a polynomial time algorithm that produces a labeling of minimum total leader length for labels of uniform size placed in fixed positions on the boundary of R.