文章摘要
秦子超,付胤胤,闫晓华.基于街景数据的平遥古城历史街区视觉景观 质量评价与优化研究[J].林业调查规划,2025,50(3):190-200
基于街景数据的平遥古城历史街区视觉景观 质量评价与优化研究
Visual Landscape Quality Evaluation and Optimization of HistoricalDistricts in Pingyao Ancient City Based on Street View Data
  
DOI：
中文关键词: 视觉景观  景观质量评价  街景数据  图像语义分割  优化策略  平遥古城历史街区
英文关键词: visual landscape  landscape quality evaluation  street view data  image semantic segmentation  optimization strategies  historical districts of Pingyao Ancient City
基金项目:2021 年山西省研究生教育改革课题“乡村振兴视域下研究生思政教育创新研究与实践”(2021YJJG083).
作者单位
秦子超 山西农业大学,山西 太谷 030801 
付胤胤 山西农业大学,山西 太谷 030801 
闫晓华 山西农业大学,山西 太谷 030801 
摘要点击次数: 43
全文下载次数: 52
中文摘要:
      历史街区因其综合的多元属性而被视为城市空间的独特组成部分,针对历史街区视觉景观评 价研究尚不完善,缺乏科学定量的评价方法。以平遥古城为研究对象,将自摄图像集作为图像深度学习载体,运用图像语义分割技术得到各客观视觉要素占比,构建历史街区视觉景观质量评价体系,采用AHP-CRITIC法确定权重;利用ArcGIS和SPSS软件,采用自然断点法、相关性分析法对数据进行分析。结果表明,平遥古城历史街区视觉质量区间为[0.247 799,0.782 619],平均值为 0.614 909 8,其中39.19%的采样点聚集在平均值附近,平遥古城历史街区视觉景观质量以中等质量为主。平遥古城历史街区视觉景观质量评价结果,空间分布呈现北区高南区低的特征。结合相关性分析得出,植被、建筑、载具、天空等客观视觉要素为平遥古城历史街区视觉景观质量的主要影响因素。针对视觉安全性、舒适性、丰富性3个方面提出视觉景观优化策略。
英文摘要:
      Historical districts are regarded as unique components of urban space due to their diverse integrated attributes. However, research on the visual landscape evaluation of historical districts remains insufficient, lacking scientifically quantitative assessment methods. This study takes Pingyao Ancient City as the research object, using self-collected image datasets as the deep learning input and applying image semantic segmentation technology to obtain the proportion of various objective visual elements. A visual landscape quality evaluation system for historical districts is constructed, with weights determined by the AHP-CRITIC method. ArcGIS and SPSS software are employed for data analysis using natural breaks classification and correlation analysis. The results show that the visual quality interval of historical districts in Pingyao Ancient City is [0.247 799, 0.782 619], with an average value of 0.6149 098. Approximately 39.19% of the sampling points cluster around the average, indicating that the visual landscape quality is primarily at a medium level. Spatially, the evaluation results reveal higher visual quality in the northern districts and lower quality in the southern districts. Correlation analysis identifies vegetation, buildings, vehicles, and the sky as the main influencing factors of visual landscape quality. Based on the findings, optimization strategies are proposed focusing on three aspects: visual safety, comfort, and richness.
查看全文   查看/发表评论  下载PDF阅读器
关闭