DEVELOPMENT AND QUALITY ASSESSMENT OF ANALYTICAL ROCK DRAWINGS

L. Hurni, T. Dahinden

Tobias Dahinden

hurni@karto.baug.ethz.ch

Rock drawing can be an important part of the map content in large scale topographical maps. In several map sheets of the Swiss National Map Series 1:25’000, rock and scree cover more than 50% of the sheet area. Especially for new mapping projects the cartographic design of a new rock plate cannot be achieved anymore due to high costs. The design of the mentioned Swiss example would require at least 2000 working hours.

The different styles of rock representations are quite manifold. The largest group is based on components like modulated fill hachures and ridge line representations, but there are also examples with accentuated relief shadings and/or contour lines with small equidistance. Therefore, in the presented work, first, a classification and an assessment of about 60 existing rock representations and their major design components was carried out. According to the frequency of the application of those components, a priority list was set up.

By applying existing and newly developed filter operations, it has then been tried to calculate the different (high priority) components from a high definition raster DEM of the Matterhorn area on the Swiss-Italian border. The different algorithms used will be discussed in detail in the paper. In order to derive a complete rock representation, components were then combined using different techniques such as transparent overlay or masking. Finally the cliff plate was combined with the remaining map elements of both the Swiss and of the Italian Topographic Map 1:25’000.

For the following assessment of the digital rock representations and their components, numerical criteria were set up. Those attributes covered for instance the geometrical and graphical quality of edges and shadings, the contrast of the rock elements, the black and white relation and the noise level created by unwanted artifacts. Furthermore, the degree of details in a hachure and the influence of the orientation angle of a generated feature were examined.

In a survey, manual and analytical rock representations were evaluated by both cartographic experts and laymen on the base of a questionnaire. The results were statistically interpreted. One of the major results was the observation of a difference of preferred rock representations between the laymen and (preconditioned?) experts. It was no surprise that the manual drawings were generally preferred to the analytical representations.

Finally, the answers of the survey were compared to the priority catalogue of the rock drawing components mentioned earlier. In the case of hachure it was found that fine grained hachures with high contrast were better suited to depict rock areas. Most of the participants of the survey preferred contour lines with, however, a rather large equidistance. Rock depictions using accentuated relief shading were not judged better, however a uniform grey tone in the rock area combined with other components were preferred. Finally, ridge line depictions were evaluated better than hachured areas.