Geospatial and Urban Data Science Research Group at the National University of Singapore
Updates from our group
Planning for active mobility satisfies many fundamental tenets of good urban design and planning. However, planning for active mobility is a complex endeavour due to numerous local, place-based factors that influence active mobility decisions. Recent advancements in urban data research have demonstrated the effectiveness of deep learning methods in evaluating active mobility potential for urban environments. At present, the incorporation of semantic information from deep learning models and street view imagery into spatio-temporal contexts remains a challenge. In particular, knowledge extraction from deep learning models remains an open question for urban planning and decision-making. Towards this issue, we propose a functional deep learning and network science workflow that employs open data from OpenStreetMap and Mapillary to assess factors affecting active mobility decisions and route planning. We demonstrate the generalisability of our analytical workflow through two case studies focusing on urban greenery in Nerima city (Japan) and urban visual complexity in Pasir Ris town (Singapore). Our results reveal clear patterns of heterogeneity in urban streetscapes and identify unevenness in street infrastructure provision based on destination types. Using this information, we propose specific areas for design intervention to improve active mobility planning. Our workflow is applicable for a diverse range of use cases making it relevant to a wide range of stakeholders, not limited to, urban researchers, policy makers and urban planners.
Street view imagery (SVI) is increasingly in competition with traditional remote sensing sources and assuming its domination in myriads of studies, mainly thanks to the omnipresence of commercial services such as Google Street View. Similar to other spatial data, SVI may be of variable quality and burdened with a variety of errors. Recently, this concern has been amplified with the rise of volunteered SVI such as Mapillary and KartaView, which – akin to other instances of Volunteered Geographic Information (VGI) – are of heterogeneous quality. However, unlike with many other forms of spatial data, there has not been much discussion about the quality of SVI datasets, let alone a standard and mechanism to assess them. Further, current spatial data quality standards are not entirely applicable to SVI due to its particularities. Following a multi-pronged method, we establish a comprehensive framework for describing and assessing the quality of SVI. We present a categorised set of 48 elements that suggest the quality of imagery and associated data such as geographic information and metadata. The framework is applicable to any source of SVI, including both commercial and crowdsourcing services. In the implementation, which we release open-source, we assess several quality elements of SVI datasets across 9 cities. The results expose varying quality of SVI and affirm the importance of the work. Given the exponential volume of studies taking advantage of SVI, but largely overlooking quality aspects, this work is a timely contribution that will benefit data providers, contributors, and users. It may also be applied on other forms of image-based VGI, and underpin establishing a formal international standard in the future. On a broader perspective, while providing an overdue definition of SVI, this work also reveals issues and open questions that impede delineating and assessing this diverse form of urban and terrestrial imagery.
3D city models are omnipresent in urban management and simulations. However, instruments for their evaluation have been limited. Furthermore, current instances are scattered worldwide and developed independently, hampering their comparison and understanding practices. While there are developed assessment frameworks in open data, such efforts are generic and not applied to geospatial data. We establish a holistic and comprehensive four-category framework ‘3D City Index’, encompassing 47 criteria to identify key properties of 3D city models, enabling their assessment and benchmarking, and suggesting usability. We evaluate 40 authoritative 3D city models and derive quantitative and qualitative insights. The framework implementation enables a comprehensive and structured understanding of the landscape of semantic 3D geospatial data, as well as doubles as an evaluated collection of open 3D city models. For example, datasets differ substantially in their characteristics, having heterogeneous properties influenced by their different purposes. There are further applications of this first endeavour to standardise the characterisation of 3D data: monitoring developments and trends in 3D city modelling, and enabling researchers and practitioners to find the most appropriate datasets for their needs. The work is designed to measure datasets continuously and can also be applied to other instances in spatial data infrastructures.
3D building models are an established instance of geospatial information in the built environment, but their acquisition remains complex and topical. Approaches to reconstruct 3D building models often require existing building information (e.g. their footprints) and data such as point clouds, which are scarce and laborious to acquire, limiting their expansion. In parallel, street view imagery (SVI) has been gaining currency, driven by the rapid expansion in coverage and advances in computer vision (CV), but it has not been used much for generating 3D city models. Traditional approaches that can use SVI for reconstruction require multiple images, while in practice, often only few street-level images provide an unobstructed view of a building. We develop the reconstruction of 3D building models from a single street view image using image-to-mesh reconstruction techniques modified from the CV domain. We regard three scenarios: (1) standalone single-view reconstruction; (2) reconstruction aided by a top view delineating the footprint; and (3) refinement of existing 3D models, i.e. we examine the use of SVI to enhance the level of detail of block (LoD1) models, which are common. The results suggest that trained models supporting (2) and (3) are able to reconstruct the overall geometry of a building, while the first scenario may derive the approximate mass of the building, useful to infer the urban form of cities. We evaluate the results by demonstrating their usefulness for volume estimation, with mean errors of less than 10% for the last two scenarios. As SVI is now available in most countries worldwide, including many regions that do not have existing footprint and/or 3D building data, our method can derive rapidly and cost-effectively the 3D urban form from SVI without requiring any existing building information. Obtaining 3D building models in regions that hitherto did not have any, may enable a number of 3D geospatial analyses locally for the first time.