AUTOMATING LINEAR TEXT PLACEMENT WITHIN DENSE FEATURE NETWORKS
暂无分享,去创建一个
In preparation for the 2000 census, the Census Bureau plans to provide maps with address information to local governments for review and update. The algorithms that have been used by the Census Bureau in the past have not been flexible enough to pl'ace this kind of text readably on small-scale maps. This paper describes a new non-interactive algorithm which has been developed at the Census Bureau. This algorithm simultaneously considers all the address range text which must be plotted in a certain area of the map and can make incremental adjustments to the location of each text item in the set in order to position it for maximum clarity before any positions are fixed. The flexibility that this approach provides has already made it possible to produce maps that display address ranges, census block numbers, and linear features and their names at a scale significantly smaller than comparable maps produced for the 1990 Census.
[1] Herbert Freeman,et al. On the Problem of Placing Names in a Geographic Map , 1987, Int. J. Pattern Recognit. Artif. Intell..
[2] Lee R. Ebinger,et al. Noninteractive Automated Names Placement for the 1990 Decennial Census , 1990 .
[3] E. Imhof. Positioning Names on Maps , 1975 .
[4] Steven Zoraster,et al. INTEGER PROGRAMMING APPLIED TO THE MAP LABEL PLACEMENT PROBLEM , 1986 .