New appoaches regarding characterisation and interpretation of the building alignments in cartographic generalisation

In this study, building alignments that can be considered a part of contextual generalisation have been investigated. Building alignments have been examined to generate information to be used in fully automatic generalisation by four different perspectives. First, a typology framework has been established to name and differentiate the building alignments one to another. Second, grouping algorithms for buildings in the urban blocks have been investigated and compared according to the building groups with different distribution characteristics. In third perspective, manually generalised building alignments have been observed to obtain constraints through the scale transitions by utilising geometric and structural measures. Finally, characterization of the alignments has been done by using newly proposed measures and a decision tree. Thus, building alignments can be classified based on Gestalt principles for the generalization purposes. As a result of these studies, new and alternative measures and rules have been proposed and useful knowledge has been generated for the automated building alignment generalisation.