Quantifying Urban Retrofitting with Large-scale Street-view Data

Keywords: Big data; Quantification; Urban Retrofitting; Spatiotemporal Analysis
Abstract Type: Paper Abstract

Abstract

Urbanization has entered a new phase characterized by urban changes and renovations occurring at a micro-scale and “under the roof”, as opposed to external modifications. These changes, known as urban retrofitting, integrate novel technologies and features into existing systems to promote sustainable and resilient urban development goals. Given the limitations of traditional methods using satellite remote sensing images in identifying urban retrofitting, novel methods need to be developed for urban retrofitting quantification and evaluation. In this paper, we propose a method using large-scale street view images to quantify the urban retrofitting process, evaluating where, when and what types of urban retrofitting have been happening. Under the 5R framework of urban retrofitting – re-inhabitation, re-building, re-transportation, re-capital, and re-greening, this paper presents a case study utilizing large-scale street view images, consisting of over 100,000 street view images and more than 87,000 image pairs, to quantify the urban retrofitting process occurring in the Mecklenburg County, North Carolina, where the city of Charlotte is located, from 2007 to 2023. The results reveal that the intensity of urban retrofitting has increased significantly since 2008. The spatial distribution of urban retrofitting follows a "Center-Periphery" pattern, with major retrofitting happening near the downtown area and the periphery of the city. However, despite the significant and growing intensity of urban retrofitting detected in the city of Charlotte, we find that socially vulnerable populations are not benefiting from the urban retrofitting processes as much as socially advantaged groups.