Automated reduction of visual complexity in small-scale relief shading

Abstract

Shaded relief derived automatically from digital elevation models differs distinctly from traditional manual shading. Particularly at small scales, many small topographic details that are present in terrain models disturb the clear portrayal of the main relief features. Automatic shading is therefore not appropriate for high-quality cartographic products. This paper proposes a new method of generalizing digital elevation models for deriving small-scale shaded relief that resembles the manual style. The procedure consists of the following raster operations: undesirable topographic details are smoothed with low-pass filters, and the main landforms, such as ridgelines and valleys, are detected by curvature coefficients. Two secondary grids are derived, one exaggerating ridgelines, the other deepening valley bottoms, and the two grids are combined according to the character of the terrain; the grid with exaggerated ridgelines is used in mountainous areas, and the grid with deepened valley bottoms in lowland areas. Finally, shaded relief is derived from the combined elevation model. Following these processing steps, only a few manual corrections are necessary to produce high-quality small-scale relief shading.