Detecting Building Alignments for Generalisation Purposes

Abstract

Generalisation is well recognised as a complex process that should trigger specific algorithms, on different types of objects in some logical or appropriate order. To guide the process (where and how to generalise) one solution is to distinguish characterisation from the generalisation process. Characterisation aims at finding and describing relevant 'working areas' that can be a part of an object or a set of objects. As a result, the choice of an appropriate algorithm(s) becomes easier and can be constrained by the detected properties of this new entity. This paper presents a method to both detect and characterise building alignments in an effort to improve the use of generalisation algorithms namely typification and displacement. The first step consists of the identification of building alignments from straight-line templates. The second step characterises these alignments to retain only those that are perceptually regular. The characterisation is based on an analysis of the spatial location of buildings as well as on the properties of the buildings that belong to the alignment in question. To evaluate the regularity of the distribution, estimators are proposed for each property.. At the end a global quality estimator of the perceptual alignment-based on the aggregation of the estimators -is proposed. This global estimator is used to retain the best building alignments that will then be carefully generalised. The results presented have been implemented in the Lamps2 GIS software.

Characterization and Detection of Building Patterns in Cartographic Data: Two Algorithms

“…According to previous work (Boffet and Rocca Serra, 2001, Christophe and Ruas, 2002, the homogeneities of general clusters can be realized using standard deviations of all the properties. In our approach, we improve this calculation by adopting the concept of coefficient of variance (CV = Std/M ean):…”

“…Some of the other patterns have been discussed by previous authors. For example, the collinear pattern by (Christophe and Ruas, 2002) and the grid pattern by (Anders, 2006).…”

“…The automated detection of visually significant building patterns is required for applications like automated map generalization, automated evaluation of generalized output, semantic enrichment of spatial databases, and spatial data mining. For example, collinear patterns have been extensively investigated (Boffet andRocca Serra, 2001, Christophe andRuas, 2002) in order to simplify and typify building groups. An approach to detect higher-level semantics like terraced house (Lüscher et al, 2009) made use of detected building alignments; the detection approach is however not widely applicable as the buildings in their case was topologically adjacent, which is not commonly the case.…”

Spatial Relations Using High Level Concepts

“…Zhang et al [16] proposed a categorization of building patterns which includes collinear, curvilinear, align-along-road, grid-like and unstructured. The detection of collinear building patterns has been extensively studied [17,18]. Zhang et al [16] proposed algorithms for detecting align-along-road and unstructured building patterns.…”

Web service approaches for providing enriched data structures to generalisation operators

“…The output of both services is just simple features. A further example of supporting geometries includes the computation of alignment lines, chaining together a group of map features such as buildings (Christophe and Ruas, 2002). The creation of inflection points or identification of local extremes for line generalisation (Plazanet et al, 1995) serves as a final example of creating supporting geometries.…”