| 注册
首页|期刊导航|四川大学学报(自然科学版)|基于街景图像的环境要素对包容度及居民属性偏好性影响分析

基于街景图像的环境要素对包容度及居民属性偏好性影响分析

刘昱 卿粼波 牛韶斐 黄建霖 赵炜

四川大学学报(自然科学版)2025,Vol.62Issue(3):619-630,12.
四川大学学报(自然科学版)2025,Vol.62Issue(3):619-630,12.DOI:10.19907/j.0490-6756.240242

基于街景图像的环境要素对包容度及居民属性偏好性影响分析

Analysis of the impact of environmental elements on inclusiveness and resident attribute preferences based on street view images

刘昱 1卿粼波 1牛韶斐 2黄建霖 1赵炜2

作者信息

  • 1. 四川大学电子信息学院,成都 610065
  • 2. 四川大学建筑与环境学院,成都 610065
  • 折叠

摘要

Abstract

Inclusiveness refers to the openness and accessibility of urban space to different attribute groups,while preferences refer to the preferences and selections of different resident attribute groups for environmen-tal elements,which together reflect the adaptability and satisfaction of cities to diverse needs.Therefore,ex-ploring the relationship between environmental elements and the degree of inclusiveness and residents'attri-bute preference is of great significance for urban planning under the perspective of"people-oriented".Previous studies on urban inclusiveness and residents'preferences are mostly based on interviews and questionnaires,which have the problems of high cost,low efficiency,and limited samples,etc.Therefore,based on com-puter vision technology,this study proposes a new quantitative index of inclusiveness from the perspective of diversity of residents'attributes,and explores the environmental factors affecting inclusiveness in the Qin-gyuan and Tongde neighborhoods of Chengdu city,as well as the impacts of the environmental factors on the preferences of residents.Specifically,the Segformer model is employed to extract semantic information per-taining to environmental elements in street view images and quantify their visual representation.Subse-quently,the Paddle Detection model is employed to extract crowd attribute information from the street view image.Based on this,an inclusiveness index is constructed and its spatial distribution is visualized.Further-more,the relationship between environmental elements and inclusiveness is analyzed using the Pearson corre-lation coefficient.At the same time,Pearson correlation coefficient is used to analyze the influence of environ-mental elements on different residents'attributes to get preferences.The results show that:(1)The overall inclusiveness value in the study area is high and shows spatial homogeneity,and the Inclusiveness value of Tongde community is higher than that of Qingyuan community and is more concentrated in the center area.The environmental elements in the study area have a significant effect on Inclusiveness.(2)There are signifi-cant Preferences for different environmental elements in the study area by different attribute groups.Based on the conclusions of the study,the following planning recommendations are proposed:the ratio of buildings,walls and sky should be coordinated;vehicles should be managed scientifically;the width and layout of roads should meet the needs of different modes of transportation;barrier-free paths should be installed and main-tained on the sidewalks;appropriate retention of vegetation.This study provides a scientific basis for optimiz-ing the planning layout and promoting the refinement of community governance in Chengdu.

关键词

街景图像/包容度/偏好性/人群属性/计算机视觉

Key words

Street view images/Inclusiveness/Preferences/Crowd attributes/Computer vision

分类

信息技术与安全科学

引用本文复制引用

刘昱,卿粼波,牛韶斐,黄建霖,赵炜..基于街景图像的环境要素对包容度及居民属性偏好性影响分析[J].四川大学学报(自然科学版),2025,62(3):619-630,12.

基金项目

成都市重点研发支撑计划(2023-YF09-00019-SN) (2023-YF09-00019-SN)

四川大学学报(自然科学版)

OA北大核心

0490-6756

访问量0
|
下载量0
段落导航相关论文