摘要
为了更好地完成历史文化名城规划编制工作,文章提出一种基于图像识别的空间品质变化评价方法。其中,利用ArcGIS工具对文化古城的街景图像进行采样,然后利用DeepLabV3+网络对街景图像的客观要素进行识别。基于以上识别结果,以机动化程度、空间围合程度、空间绿视率和可步行道路占比作为空间品质评价指标,进而对北京某区域进行评价。最后,基于评价结果,提出历史文化名城空间品质提升策略,包括对街道进行常态化监测与评估、控制机动化水平过度增长、提升可步行街道占比、增加舒适开敞的休闲区域、提高街道的绿视率等提升空间品质的措施。
In order to make a better planning of famous historical and cultural cities,a spatial quality change evaluation method based on image recognition is proposed.Among them,ArcGIS tool is used to sample the street images of the ancient cultural cities,and then DeepLabV3+network is used to identify the objective elements of street view images.Based on the above identification results,the degree of motorization,the degree of spatial enclosure,the rate of spatial green vision and the proportion of walkable roads were taken as the evaluation indexes of spatial quality,and then an area in Beijing was evaluated.Finally,based on the evaluation results,the strategies for improving the spatial quality of famous historical and cultural cities were proposed,including regular street monitoring and evaluation,controlling excessive growth of motorized level,increasing the proportion of walkable streets,increasing comfortable and open leisure areas,improving the green visual rate of streets,and other measures.
出处
《城市建筑》
2024年第6期8-11,共4页
Urbanism and Architecture
关键词
历史文化名城
城市空间品质
图像语义分割
评价体系
品质提升
famous historical and cultural city
urban spatial quality
image semantic segmentation
evaluation system
improvement of quality
作者简介
曹诚为(1982-),女,副高级工程师,研究方向:城市规划及国土空间规划。
