基于绿视率和NDVI的城市街道景观分析与优化研究  

nalysis and Optimization of Urban Street Landscape Based on GVI and NDVI

作　　者：苏雷 陈伟峰 李俊英 周燕[3,4] 樊磊 SU Lei;CHEN Wei-feng;LI Jun-ying;ZHOU Yan;FAN Lei(College of Art and Design,Zhongshan Institute,University of Electronic Science and Technology,Zhongshan 528402,Guangdong,China;Architectural and Civil Engineering College,Huizhou University,Huizhou 516007,Guangdong,China;College of Forestry,Shenyang Agricultural University,Shenyang 110161,Liaoning,China;Key Laboratory of Northern Landscape Plants and Regional Landscape of Liaoning Province,Shenyang 110161,Liaoning,China)

机构地区：[1]电子科技大学中山学院艺术设计学院,广东中山528402 [2]惠州学院建筑与土木工程学院,广东惠州516007 [3]沈阳农业大学林学院,辽宁沈阳110161 [4]辽宁省北方园林植物与地域景观重点实验室,辽宁沈阳110161

出　　处：《西北林学院学报》2024年第2期256-264,共9页Journal of Northwest Forestry University

基　　金：国家自然科学基金项目(31670703);广东省科技计划项目(2020-409);广东省教育厅特色创新项目(420N50)。

摘　　要：街道景观空间对市民健康和城市风貌具有重要影响。既往研究中常以归一化植被指数(NDVI)和绿视率(GVI)来分别代表二维和三维的绿色指标,但对二者的指标相关性研究甚少。采用基于深度学习的图像语义分割方法分析百度街景计算代表性街道的GVI,利用GF-1卫星数据计算NDVI,比较分析城市街道的GVI和NDVI指标特征及相关性。结果表明,1)中山市中心城区各代表街道GVI指标参差不齐,从8.06%到36.00%,其中石岐街道兴中道GVI最高;2)各街道观测点的NDVI均值随着缓冲区尺度的增加也随之呈现出不同变化,NDVI均值具有强烈的尺度敏感性;3)50 m GVI和DNVI均值的皮尔逊相关系数最高,达到0.832。在此基础上分析街道景观存在的不足并给出优化建议,为城市街景评估、空间优化、景观提升提供参考。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 often 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 representative 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 increase of buffer scale,and the mean value of NDVI had a strong scale sensitivity.3)The Pearson correlation 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 

分 类 号：S731.1[农业科学—林学]