Theory Of The Cartographic Line Revisited/ Implications For Automated Generalization

Abstract

‘The Theory of a Cartographic Line’ (Peucker 1975) describes width as being the essential characteristic of a cartographic line. Digital representations have tended to ignore this basic attribute and in the context of generalization the oversight is detrimental. The theory claims that a set of enclosing bands captures the cartographic character of width and supports generalization. The Douglas algorithm, still one of the most commonly used algorithms for generalizing digital representations, uses this model. Work of a Polish mathematician Perkal, provides the basis for another model of cartographic line width and a different generalization technique. This paper examines how effectively both models capture cartographic line width and succeed in producing generalized results, particularly for larger scale reductions. The two techniques are assessed by their ability to satisfy two objectives: capturing the essential and recognizable characteristics of geographic features and creating representations which can be legibly displayed at smaller scales. The paper compares the behavior of the two methods as applied to digital coastline data.