AUTOMATED CARTOGRAPHIC TEXT PLACEMENT
Lin Li, Yadong Shu
Cartographic annotation is composed of nature language, and is one part of Map Language. Cartographic label is the supplement to map symbol. It makes Map more readable, interpretable. It becomes the most important facility for transmission of map-information. The most time-consuming task of producing a map-after data collection- is labeling the point, line and area features. With the development of cartography, It is an urgent task to find a best method that can place the text quickly , reasonably and automatically.
There are three types of features(point, line, area)in the label placement on the relief map. As to the point and line features, the automatization methods of text placement are relatively mature. The automatization method of point feature text placement is mainly based on the eight-direction. The automatization method of line feature text placement is more complex. First, builds the buffer of the object line, and then gets the text place along the buffer line. As to the area features ,most of the studies pay more attention to the algorithms ,such as the skeleton and center of polygon, and these algorithms are the same with only one polygon(The label is interior of the polygon.). So it is not fit to the dispersed area feature, for instance, the resident area(The label is exterior of polygon). It is rarely to see any study on automatization method about the dispersed area feature.
This paper mainly studies the label methods of point, line and area. And it simply concludes and implements the label methods of point and line. As to the problem of wrong label direction lain in labeling line features, a solution based on the theory of vision is put forward. According to the needs of common placement, this paper discusses the area labeling from the two ways----labeling in area and out area.
For traditional labeling in area, based on the Delaunay triangulation the author first gets the skeleton line and center of polygon. And then, after getting the optimal position through the method of auto- adaptive, the annotation is placed along the skeleton line.
For the situation of labeling out area, the author first obtains the convex hulls of the dispersed polygons, and then gets the candidate placement from the sixteen directions in the convex hull. But the best initial text place mustn't be the candidate place. The best initial position is got from the self-adjust method based on auto- adaptation.
At last, the author apply this annotation method to the mapping of our national 1:5,000 scale topographic maps of china, which has achieved good effects and embodies the cartographer’s rules, greatly improves the degree of automation of topographic annotation and accelerates the speed of annotation.