打印

作者

Waishan Qiu;Wenjing Li;Xun Liu;Ziye Zhang;Xiaojiang Li;Xiaokai Huang

来源

CITIES,Vol.133,Issue1,Article 104037

语言

英文

关键字

Street view image;Subjective and objective perceptions;Housing prices;Machine learning;Computer vision

作者单位

Department of City and Regional Planning, Cornell University, Ithaca, NY 14850, USA;Center for Spatial Information Science, The University of Tokyo, Tokyo 113-8654, Japan;School of Architecture, University of Virginia, Charlottesville, VA 22904, USA;Paul and Marcia Wythes Center on Contemporary China, Princeton University, Princeton, NJ 08544, USA;Department of Geography and Urban Studies, Temple University, Philadelphia, PA 19122, USA;Graduate School of Design, Harvard University, Cambridge, MA 02138, USA;Department of City and Regional Planning, Cornell University, Ithaca, NY 14850, USA;Center for Spatial Information Science, The University of Tokyo, Tokyo 113-8654, Japan;School of Architecture, University of Virginia, Charlottesville, VA 22904, USA;Paul and Marcia Wythes Center on Contemporary China, Princeton University, Princeton, NJ 08544, USA;Department of Geography and Urban Studies, Temple University, Philadelphia, PA 19122, USA;Graduate School of Design, Harvard University, Cambridge, MA 02138, USA

摘要

Recently, housing prices studies emerged to use street view imagery to infer the marginal price of streetscape using hedonic price models (HPM). Within this trend, most studies took an objective approach by extracting streetscape pixels to directly proxy human perceptions. Less studies addressed the strength of the subjective perceptions measured from visual surveys. We argue that human perception is a subjective and subtle sensory process that cannot be fully captured using objective indicators. We hypothesize that the subjective approach could provide stronger predictive power to housing prices. Taking Shanghai as a case study, we extend a spatial HPM with both subjectively-measured and objectively-derived perceptions of six urban design qualities. Accounting for the spatial dependence, subjective scores collectively explained equal price variance, while all their individual signs were opposite to the objective counterparts. Moreover, most subjective scores individually exhibited stronger strength, confirming our hypothesis that perceptions are better captured using subjective measures. For ambiguous concepts such as imageability, the subjective framework exhibits better performance, while the objective constructs are more effective to represent straightforward perceptions like greenness and enclosure. This study contributes to the studies on the coherence and divergence of subjective and objective measures of perceptions and environmental qualities. It enriches the literature on the economic values of streetscapes and provides guidance on the indicator selection for assessing, designing, and managing street environments.