Multiscale analysis of the influence of street built environment on crime occurrence using street-view images. (October 2022)

• Record Type: Journal Article
• Authors: HE, Zhanjun; Wang, Zhipeng; Xie, Zhong; Wu, Liang; Chen, Zhanlong
• Is Part Of: Computers, environment and urban systems. Volume 97(2023)
• Publication Date: 2022-10
• Subjects: Street built environment -- Crime prevention -- Google street view -- Multiscale analysis -- Multiscale geographically weighted regression; City planning -- Data processing -- Periodicals; Regional planning -- Data processing -- Periodicals; 303.4834
• DOI: 10.1016/j.compenvurbsys.2022.101865
• Languages: English
• ISSNs: 0198-9715

Abstract

Abstract: Assessing the effect of street built environment on crime occurrence is a hot research subject in environmental criminology, which also plays an important role in crime prevention or even urban planning. Recent development in emerging geotagged big data (e.g., the street-view images) makes it possible to quantify the influence of street built environment on crime. However, previous studies have often neglected the multiscale problem in exploring the association between environmental features and crime occurrence. Therefore, in this study, a multiscale analysis method was proposed to quantitatively study the influence of street built environment on crime occurrence using street-view images. Firstly, inspired by the theory of crime prevention through environmental design, we established a multiscale descriptive framework for environmental features with simultaneous consideration of the physical features and scene perception of street built environment. Then, a multiscale geographically weighted regression model was used to explore the spatial scale of influence for different streetscape features on crime occurrence. Experimental results indicated that the proposed method could reflect the difference of the spatial scale of various environmental features on crime, thereby uncovering the association between environmental features and crime occurrence with better accuracy. This study may enrich the theory in environmental criminology, and it provides useful insights for crime prevention through urban streetscape design. Highlights: A multiscale descriptive framework for environmental features is established to quantify street built environment. The MGWR model was used to explore the spatial scale of influence for different streetscape features on crime occurrence. The proposed multiscale analysis method well connects the criminological theory (CPTED) and geospatial big data.