Assessing equity in visual landscape quality toward a “15-minute living circle” in Binjiang, Hangzhou: using deep learning and image recognition

In the process of promoting the construction of livable cities, visual landscape quality (LVQ) has gradually become an important indicator of urban livability. However, there are significant imbalances in visual landscape quality among different communities due to differences in economic, policy and other factors, and this variability has triggered public concern about its fairness. An in-depth assessment of the equity of visual landscape quality can help provide strong theoretical support and practical guidance for the scientific allocation of landscape resources. This study aims to assess the fairness of the visual landscape quality of the community living circle, for which a fairness assessment framework is constructed, which integrates the visual landscape quality, deep learning and image recognition, spatial statistical analysis, and community house price big data, to conduct a community-level visual landscape for the “15-minute living circle” in the Binjiang District of Hangzhou as an example of the quality fairness assessment. The results show that there is a obvious inequity in visual landscape quality between residential neighbourhoods, with the Riverside South Strategic and Central Enhancement Districts having higher visual landscape quality than the Riverside Quality District, and that there is a spatially negative correlation between community housing prices and visual landscape quality in this study area. The findings provide relevant insights for the scientific allocation of landscape resources and the elimination of inequity in visual landscape quality in urban areas.