Elvas Duarte de Almeida, J-P.; (2008) A graph-based technique for analysis and visualisation of higher order urban topology. Doctoral thesis , University of London.

Text U591208.pdf Download (39MB)

Abstract

Analysis of spatial phenomena is a time consuming and laborious task in several fields of the Geomatics world. Automation of these tasks is especially needed in areas such as geographical information science (GISc). Carrying out those tasks in the context of an urban scene is particularly challenging given the complex spatial pattern of its elements. The starting point in this study is unstructured data, and hence no prior knowledge of the spatial entities is assumed. The aim of translating this data into more meaningful homogeneous regions can be achieved by grouping geographic structures within the initial collection of objects according to their spatial arrangement. The techniques applied to achieve this are those of graph theory, applied to urban topology analysis. For the identification of meaningful structures a graph-based system was developed comprising, in particular, a procedure, the containment-first search, based on the breadth-first search algorithm for graph traversal which, by considering the spatial objects' adjacencies, analyzes and interprets their spatial arrangement in terms of the topological relationship of containment. Different LiDAR as well as photogrammetric datasets have been used as an example scenario to test this system. Another aspect is the visualisation of the urban topology. Visual inspections of interim results of the graph analysis process can often reveal patterns not discernable by current automated techniques. An interactive tool was developed and implemented in Visual Basic for Applications (VBA) utilising ESRFs ArcObjects, in ArcMap (ArcGIS). The ultimate goal was the dynamic display of the original map according to the results of the spatial topology analysis. Future work will entail further clustering of the identified containment units into homogeneous regions. After the delineation of cluster shapes, an analysis process will have to be accomplished, either by pattern recognition or interpretation procedures, for the retrieval of higher-level information, for example related to land-us.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item