SHAPE ANALYSIS OF MULTI-SCALE BUILDING FEATURES

Abstract

Buildings can generally be represented as individual polygons until 1:100 000 in topographic maps. As scale decreases, the level of detail of the features is reduced with generalization to preserve the readability on map. In this context, shape analysis is important because it can be used in generalization operator and parameter selection, generalization quality evaluation as well as scale inference for VGI and remotely sensed data. Shape analysis of polygons are often carried out with various shape indices. A shape index usually measures a specific shape characteristic of a polygon. In this paper, it is aimed at investigating the response of different shape indices to the level of detail variation in building features depending on scale. For this purpose, five shape indices were used and evaluated for measuring shape complexity of individual building features at six different scales ranging from 1:1 000 to 1:100 000.