Simple luminosity normalization of greenness, yellowness and redness/greenness for comparison of leaf spectral profiles in multi-temporally acquired remote sensing images
Simple luminosity normalization of greenness, yellowness and redness/greenness for comparison of leaf spectral profiles in multi-temporally acquired remote sensing images
复制标题
绿度、黄度和红/绿度的简单光度归一化，用于比较多时相获取的遥感图像中的叶片光谱剖面
DOI：
10.1007/s12038-012-9241-3
复制
发表时间：
2012
影响因子：
2.9
通讯作者：
R. Doi
中科院分区：
文献类型：
--
作者：
R. Doi
通过遥感观测叶色（光谱剖面）是查明叶色异常空间分布模式的一种有效方法，可据此采取适当的植物管理措施。然而，由于遥感图像的亮度随采集时间而变化，在观测多时相遥感图像中的叶片光谱轮廓时，必须考虑亮度的变化。这项研究确定了一个简单的光度归一化技术，使树叶的颜色进行比较，随着时间的推移在遥感图像。利用不同时间点获取的遥感图像对曼谷和东京的各种不变屋顶进行光谱剖面分析时，绿色和黄色（绿色+红色）的强度值与光度呈现出较强的线性关系（R2 > 0.926）。描述绿色或黄色强度的公式的系数和常数或系数的值在单个曼谷站点和两个东京站点之间具有可比性，表明该技术的普遍适用性。对于单屋顶，绿色、黄色和红色/绿色的变异系数值为16%或更小（n = 6 - 11），表明精度不低于公认的遥感测量，如归一化差异植被指数。在获得上述线性关系之后，对原始强度值进行归一化，并对东京万年青和落叶树种的冠层光谱轮廓进行时间比较，以突出冠层光谱轮廓的变化。本文讨论了这种技术的未来方面。
Observation of leaf colour (spectral profiles) through remote sensing is an effective method of identifying the spatial distribution patterns of abnormalities in leaf colour, which enables appropriate plant management measures to be taken. However, because the brightness of remote sensing images varies with acquisition time, in the observation of leaf spectral profiles in multi-temporally acquired remote sensing images, changes in brightness must be taken into account. This study identified a simple luminosity normalization technique that enables leaf colours to be compared in remote sensing images over time. The intensity values of green and yellow (green + red) exhibited strong linear relationships with luminosity (R2 > 0.926) when various invariant rooftops in Bangkok or Tokyo were spectral-profiled using remote sensing images acquired at different time points. The values of the coefficient and constant or the coefficient of the formulae describing the intensity of green or yellow were comparable among the single Bangkok site and the two Tokyo sites, indicating the technique’s general applicability. For single rooftops, the values of the coefficient of variation for green, yellow, and red/green were 16% or less (n = 6 − 11), indicating an accuracy not less than those of well-established remote sensing measures such as the normalized difference vegetation index. After obtaining the above linear relationships, raw intensity values were normalized and a temporal comparison of the spectral profiles of the canopies of evergreen and deciduous tree species in Tokyo was made to highlight the changes in the canopies’ spectral profiles. Future aspects of this technique are discussed herein.
使用光学扫描仪量化植物诊断中的叶绿度和叶光谱轮廓
影响因子：
1.1
作者：
Doi;R.
通讯作者：
R.