基于街景图像的不同类型街道景观评价研究

Landscape Evaluation of Different Types of Streets Based on Street View Images

街道不同的功能导向决定其在街道要素、建筑风格、交通需求等方面存在差异,进而影响街道景观质量。以商业型、生活型、景观休闲型3种类型街道为研究对象,以百度街景图片为主要数据来源,使用DeepLabv3+深度卷积神经网络模型、核密度分析等方法对3种不同类型街道的景观质量进行分析评价。研究结果表明:1)商业型街道更关注街道机动通行指标与色彩丰富指标,对天空开阔程度与空间围合程度重视不够;2)生活型街道重视生活氛围,机动干扰程度低,空间围合程度适宜,但天空开阔程度低、色彩丰富程度欠缺;3)景观休闲型街道偏向旅游体验,天空开阔程度、空间围合程度与色彩丰富度分布均衡,但部分街道机动车干扰程度高、绿化水平有待提升;针对上述问题提出相应的景观质量提升策略。以期为今后城市街道景观研究提供新的思路。

Differences in street elements, architectural styles and traffic demands are determined by the different functions of the street, which will affect the quality of street landscape.Dongjiekou Block, Baimahe Block and Wushanlu Block are respectively regarded as commercial street, life street and landscape leisure street in Fuzhou.Baidu Street View images were used as the main data, DeepLabv3+ model and kernel density analysis were used to compare and analyze the landscape visual quality of three kinds of streets.The results show that: 1) Commercial streets pay more attention to the vehicle interference index and color richness index, and pay less attention to the degree of sky visibility index and space enclosure index;2) Living streets pay more attention to living atmosphere, with low vehicle interference index and appropriate space enclosure index, but lack of sky visibility and color richness;3)The landscape leisure streets are more inclined to tourism experience, and the degree of sky openness, space enclosure and color richness are evenly distributed. However, some streets have high vehicle interference and the greening level needs to be improved.In view of the above problems, the corresponding strategies to improve landscape visual quality are proposed.In order to provide a new idea for future urban street landscape research.