Measuring the Cityscape: A Pipeline from Street-Level Capture to Urban Quantification. Issue 1 (1st September 2022)

• Record Type: Journal Article
• Authors: Ward, W; Dai, M; Arbabi, H; Sun, Y; Tingley, D; Mayfield, M
• Is Part Of: IOP conference series. Volume 1078:Issue 1(2022)
• Publication Date: 2022-09-01
• Subjects: building stock -- 3-D modelling -- street-level capture -- computer vision; Earth sciences -- Periodicals; Environmental sciences -- Congresses; Environmental sciences -- Periodicals; 550.5
• DOI: 10.1088/1755-1315/1078/1/012036
• Languages: English
• ISSNs: 1755-1307

Abstract

Abstract: Any solution to achieving climate targets must be performed at scale. Data driven methods allow expert modelling to be emulated over a large scope. In the UK, there are nearly 30 million residential properties, contributing to over 30% of the national energy consumption. As part of the UK Government's requirement to meet net-zero emissions by 2050, retrofitting residential buildings forms a significant part of the national strategy. This work addresses the problem of identifying, characterising and quantifying urban features at scale. A pipeline incorporating photogrammetry, automatic labelling using machine learning, and 3-D geometry has been developed to automatically reconstruct and extract dimensional and spatial features of a building from street-level mobile sensing.