基于街景图像和计算机视觉技术的可骑行性评价研究动态
Academic Dynamics on Bikeability Assessment Based on Street View Imagery and Computer Vi-sion
杨晰涵1
作者信息
- 1. 同济大学交通运输工程学院,上海 201804
- 折叠
摘要
选取来自国际学术期刊的论文,以概述形式对城市交通理论方法、实证分析等学术研究成果进行总结性介绍,旨在增强城市交通业界和学界对国际学术动向和研究热点的关注,促进学术交流.本论文对街景图像和计算机视觉技术在可骑行性评价中的应用进行探索,构建包含5类34个指标的可骑行性指数,并在新加坡和东京进行实证研究.结果表明,将街景图像和计算机视觉技术相结合能够有效评价可骑行性,且街景图像指标显著优于非街景图像指标,然而兴趣点数量、土地混合利用指数、道路坡度、空气质量指数等非街景图像指标的作用亦不容忽视.
Abstract
A review of selected articles from international academic journals is presented to summarize re-search findings,theoretical approaches,and empirical analysis of urban transportation.The aim is to en-hance the communication between industrial and academic fields of urban transportation,highlight interna-tional research focuses,and promote academic exchange.This paper discusses the application of street view imagery(SVI)and computer vision technology in bikeability assessment.The paper develops a bike-ability index with 34 indicators in five categories and implements it in Singapore and Tokyo.The findings demonstrate that integrating street view imagery with computer vision technology can effectively assess bikeability and SVI indicators are significantly effective than the non-SVI indicators,while the roles of the number of Point of Interest,land use mix index,road grade,Air Quality Index,and other non-SVI indica-tors should not be overlooked.
关键词
可骑行性/街景图像/计算机视觉Key words
bikeability/street view imagery/computer vision引用本文复制引用
出版年
2024
NETL
NSTL
万方数据