Modelling Urban Transformations: A Spatio-Temporal Analysis of Building Changes in the Conservation Area of Utrecht

Abstract

This thesis studies transformation of urban environments, taking the conservation area of Utrecht as a case study. It aims to explore methods for vectorization, object matching and change detection of building data across different time periods. By studying these three processes together, this study aims to develop a holistic approach for linking historical and modern building data, bridging the gap between past and present datasets and offering a pathway for more integrated analyses of the built environment. This research found that combining manual and automatic vectorization provides the most solid foundation for extracting buildings from historical cadastral maps. For object matching, a 1:1 match using building centroids, supplemented by ‘contains’ and ‘within’ spatial relationships, proved most effective in matching BGT Pand building versions. A rule-based change detection method, refined through 1:m spatial joins, was the most reliable for identifying building changes in the BGT Pand dataset between 2016 and 2024. It can be concluded that relatively many changes have occurred in the conservation area since the beginning of the BGT Pand version history in 2016, despite being an area that has strict instructions for the modification of buildings.