文献类型：期刊文献

中文题名：深圳市10种彩叶植物高光谱特征分析

英文题名：An analysis of the hyperspectral characteristics in multiple kinds of garden color plants in Shenzhen city

作者：刘东蔚[1] 孙冰[1] 陈勇[1] 蔡刚[1]

第一作者：刘东蔚

机构：[1]中国林业科学研究院热带林业研究所

年份：2014

卷号：38

期号：2

起止页码：111-116

中文期刊名：南京林业大学学报：自然科学版

收录：CSTPCD;;北大核心:【北大核心2011】；CSCD:【CSCD2013_2014】；

基金：中央级公益性科研院所基本科研业务费专项资金项目(RITFKYYW2010-12);深圳市技术研究开发计划(JSA200904031813A)

语种：中文

中文关键词：彩叶植物;高光谱特征;反射面积值

外文关键词：color plant;hyperspectral characteristic;reflection area value

分类号：S718.43

摘要：地物光谱特征是遥感应用研究的基本依据。彩叶植物对于建设色彩丰富、植物季相鲜明的园林城市有重要作用。以深圳的金叶榕（ Ficus microcaba cv． Golden Leave）、朱焦（ Cordyline terminalis）、红苋草（ Alternanthera bettzickiana）等10种彩叶植物为研究对象，应用Fieldspec3便携式地物波谱仪进行了光谱数据的采集和处理。结果表明：红叶系和紫叶系植物反射光谱在波长600～650 nm间有一个“红峰”，而在500～550 nm范围内紫色系植物光谱反射区面积大于红色系植物；黄色系植物叶片反射光谱的表现明显特征的波长区为490～650 nm，其光谱反射率在490 nm处急剧上升，并在550～650 nm区间保持0．6以上。3种色系彩叶植物叶片500～550 nm光谱的反射面积值（ Sλ500～550）差异极显著，聚类分析结果与实际色系分类一致，故Sλ500～550可以作为指示叶片色彩色系的特征值。
The spectral feature of ground objects is the fundamental basis of remote sensing application research. The color plants application is widely concerned in the city landscape, and which plays an important role in constructing a color plentiful and distinct seasonal changing garden city. Hyperspectral data of the samples were acquired by using ASD Fieldspec 3 spectroradiometer,the samples were 10 species of color plants, including Ficus microcaba cv.Golden Leave, Alternanthera bettzickiana, Cordyline terminalis and so on in Shenzhen. The results of analysis showed that the reflection spectrum of red-leaf plant and purple leaf plant had a“red peak” value in the 600-650 nm, while in the 500-550 nm range the spectral reflection area of purple leaf plant was significantly greater than red-leaf plant. For the yellow leaf plant, the distinct waveband of reflection spectrum was 490-650 nm, and the spectral reflectance increased sharply at 490 nm, then the level kept more than 60% in the 550-650 nm range;the variance analysis showed that the spectral re-flection area value of 500-550 nm was significantly different. Moreover, the cluster analysis result was consistent with actual color classification, so the value of Sλ500-550 could be used as a color distinct value of plant color.