西北林学院学报2024，Vol.39Issue(2)：256-264,9.DOI:10.3969/j.issn.1001-7461.2024.02.32
基于绿视率和NDVI的城市街道景观分析与优化研究
Analysis and Optimization of Urban Street Landscape Based on GVI and NDVI
摘要
Abstract
Street landscape space has an important impact on citizens'health and urban style.In previous studies reported,the normalized difference vegetation index(NDVI)and green visual index(GVI)are of-ten used to represent two-dimensional and three-dimensional green indicators,respectively.However,little research has been conducted on the correlation between the two indicators.To analyze the characteristics and correlation between GVI and NDVI of urban streets,this study used a deep learning method based on the image semantic segmentation to analyze Baidu Street View and to calculate the GVI of representative streets.GF-1 satellite data was used to calculate the NDVI.The results showed that 1)the GVI of repre-sentative streets in the central urban area of Zhongshan varied from 8.06％to 36.00％,with Shiqi Street having the highest GVI.2)The mean value of NDVI of each street showed different changes with the in-crease of buffer scale,and the mean value of NDVI had a strong scale sensitivity.3)The Pearson correla-tion coefficient of the 50 meter GVI and NDVI averages was the highest,reaching 0.832.On this basis,the shortcomings of street landscape were analyzed and optimization suggestions were given,which can provide references for the urban street landscape evaluation,spatial optimization,and landscape improvement.关键词
绿视率(GVI)/街景地图/归一化植被指数(NDVI)/深度学习/景观优化Key words
green visual index(GVI)/street view map/normalized difference vegetation index(NDVI)/deep learning/landscape optimization分类
农业科技引用本文复制引用
苏雷,陈伟峰,李俊英,周燕,樊磊..基于绿视率和NDVI的城市街道景观分析与优化研究[J].西北林学院学报,2024,39(2):256-264,9.基金项目
国家自然科学基金项目(31670703) （31670703）
广东省科技计划项目(2020-409) （2020-409）
广东省教育厅特色创新项目(420N50). （420N50）
