As a process to find proper positions for annotations, annotation placing has been regarded as a difficult problem in automatic map making. Following accepted cartographic conventions, the objective of a good label placement is to display the geographic position of features with their corresponding text in a clear and harmonious fashion, and having no overlap most important. In this paper, we think that annotation can have two categories by shape. The first is named point locating annotation; the second is named line locating annotation. Both of them can be seen from a combinatorial optimization point of view. Our research first evaluates the heuristic Tabu Search (TS), and then applies this method to cartographic label placement. Lastly, based on an experiment made on 1/250000 topographic data of China, TS has proven to be an efficient choice, with the best performance in quality. And results show that labels of cartographic features almost have no overlap.
[1]
Fred W. Glover,et al.
Tabu Search - Part I
,
1989,
INFORMS J. Comput..
[2]
Gilberto Câmara,et al.
Spring: integrating remote sensing and gis by object-oriented data modelling
,
1996,
Comput. Graph..
[3]
Stuart M. Shieber,et al.
Placing Text Labels on Maps and Diagrams
,
1994,
Graphics Gems.
[4]
Joe Marks,et al.
An empirical study of algorithms for point-feature label placement
,
1995,
TOGS.
[5]
Fred Glover,et al.
Tabu Search - Part II
,
1989,
INFORMS J. Comput..
[6]
Stephen A. Hirsch,et al.
An Algorithm for Automatic Name Placement Around Point Data
,
1982
.
[7]
Fred Glover,et al.
Tabu Search: A Tutorial
,
1990
.
[8]
C. Reeves.
Modern heuristic techniques for combinatorial problems
,
1993
.