刚竹毒蛾危害下的毛竹叶片光谱特征波长研究

引用本文

黄旭影, 许章华, 林璐, 刘健, 钟兆全, 周华康. 刚竹毒蛾危害下的毛竹叶片光谱特征波长研究. 光谱学与光谱分析, 2018,38(12): 3829-3838
HUANG Xu-ying, XU Zhang-hua, LIN Lu, LIU Jian, ZHONG Zhao-quan, ZHOU Hua-kang. Spectral Characteristic Wavelengths of Moso Bamboo Leaves Damaged by Pantana Phyllostachysae Chao . Spectroscopy and Spectral Analysis, 2018,38(12): 3829-3838. 复制到剪切板

Doi: 10.3964/j.issn.1000-0593(2018)12-3829-10
Permissions

光谱学与光谱分析所有

刚竹毒蛾危害下的毛竹叶片光谱特征波长研究

黄旭影¹, 许章华^1,^2,^3,^4,^*, 林璐¹, 刘健³, 钟兆全⁵, 周华康⁶

1. 福州大学环境与资源学院, 福建福州 350116

2. 福州大学信息与通信工程博士后科研流动站, 福建福州 350116

3. 福建省资源环境监测与可持续经营利用重点实验室, 福建三明 365004

4. 空间数据挖掘与信息共享教育部重点实验室, 福建福州 350002

5. 福建省顺昌县国有林场, 福建南平 353200

6. 福建省南平市延平区林业局, 福建南平 363000

*通讯联系人 e-mail: fafuxzh@163.com

作者简介: 黄旭影, 1993年生, 福州大学环境与资源学院硕士研究生 e-mail: 332153458@qq.com

收稿日期: 2017-03-23 接受日期: 2017-10-28

基金项目: 国家自然科学基金项目(41501361), 中国博士后面上基金项目(2018M630728), 福建省资源环境监测与可持续经营利用重点实验室开放基金项目(ZD1403), 福州大学人才基金项目(XRC-1345), 福州大学国家级本科生科研训练计划项目(201710386020)资助

摘要

旨在获取刚竹毒蛾危害下的毛竹叶片光谱特征波长, 以助于该虫害的有效、准确识别。将于福建省顺昌县实测的105条高光谱数据随机划分为实验组(71条)和验证组(34条)。基于实验组数据, 利用单因素方差分析获取健康、轻度危害、中度危害、重度危害等虫害等级间具有极显著差异的波长; 结合常用遥感卫星的波段设置对上述波长进行筛选, 采用欧式距离、相关系数及光谱角匹配等3种方法判定其虫害判别能力, 获取特征波长, 并引入验证组样本对其予以验证。结果表明: (1)受害叶片的光谱反射率明显低于健康叶片, 虫害等级越高, 其反射率越低; (2)受害叶片的光谱特征变化较大, 随着虫害等级的上升, 其光谱曲线中的“绿峰”及“红谷”趋于消失, “红边”斜率逐渐减小; (3)确定原始光谱703.43~898.56 nm及一阶微分光谱497.68~540.72, 554.53~585.25和596.24~618.23 nm为刚竹毒蛾危害下的毛竹叶片光谱特征波长, 其对该虫害具有较强的判别能力。该研究从叶片尺度剖析了寄主对刚竹毒蛾的响应机理, 是“地-天”耦合的理论基础, 可为虫害遥感监测技术体系的建立提供重要依据。

关键词: 刚竹毒蛾; 毛竹叶片; 特征波长; 光谱微分; 虫害判别能力; 顺昌县

中图分类号:TP79 文献标识码:A

Spectral Characteristic Wavelengths of Moso Bamboo Leaves Damaged by Pantana Phyllostachysae Chao

HUANG Xu-ying¹, XU Zhang-hua^1,^2,^3,^4,^*, LIN Lu¹, LIU Jian³, ZHONG Zhao-quan⁵, ZHOU Hua-kang⁶

1. College of Environment and Resources, Fuzhou University, Fuzhou 350116, China

2. Postdoctoral Research Station of Information and Communication Engineering, Fuzhou University, Fuzhou 350116, China

3. Fujian Provincial Key Laboratory of Resources and Environment Monitoring & Sustainable Management and Utilization, Sanming 365004, China;

4. Key Lab of Spatial Data Mining & Information Sharing, Ministry of Education, Fuzhou 350002, China;

5. State-owned Forest Farm of Shunchang County, Nanping 353200, China

6. Yanping District Forestry Bureau of Nanping, Nanping 353000, China

Abstract

The paper aims to obtain the characteristic wavelengths of moso bamboo leaves damaged by Pantana phyllostachysae chao, with which the pest can be identified effectively and accurately. 105 hyperspectral data collected in Shunchang County, Fujian Province were randomly divided into two groups, i.e. the experimental group (71) and verificantion group (34). Selecting wavelengths which were highly significant differences between different pest levels group by One-way ANOVA. The wavelengths were screened by combining the wave band of the commonly used remote sensing satellite. The ability to discriminate between the pests of the selected wavelengths was analyzed by the Euclidean distance method, spectral angle mapping method and correlation coefficient method. According to the analysis results, the characteristic wavelengths were obtained and verified. The results showed that: (1) the spectral reflectance of moso bamboo leaves damaged by P. phyllostachysae were significantly lower than that of healthy leaves, and the higher the pest level is, the lower the reflectance will be; (2) with the increase in pest level, the spectral reflectance curves’ “green peak” and “red balley” of Pinus massoniana gradually disappeared, and the red edge was leveled; (3) the characteristic wavelengths of 703.43~898.56 nm (original spectrum) and 497.68~540.72, 554.53~585.25, 596.24~618.23 nm (first derivative spectrum) were determined, which had good response ability at different pest levels. Our findings will not only provide the theoretical guidliances for “ground-space” coupling, but also provide important basis for establishing the system of pest remote sensing monitoring technology.

Key words: Pantana phyllostachysae Chao; Moso bamboo leaves; Characteristic wavelengths; Derivative spectrum; Shunchang County

文章图片

引言

我国是全球竹林分布最广的国家, 竹种数量、竹林面积、竹产品产量等均居世界首位。在诸多竹种中, 毛竹占绝对优势, 对竹林经济、社会与生态效益起关键作用。刚竹毒蛾、毛竹叶螨、竹蝗、枯梢病等病虫害是毛竹林面临的主要威胁, 其多呈由点及面扩散特征, 一旦蔓延, 防治困难, 给国家与林农带来极大的经济损失。随着林业科技的进步尤其是林业遥感技术水平的不断提升, 以及生态文明战略的不断推进, 森林虫害遥感研究重新成为热点^{[1, 2, 3]}。

刚竹毒蛾是毛竹最主要的食叶害虫, 受害毛竹竹体瘦小, 竹节内积水, 产笋量急剧下降, 大发生时将竹叶食尽, 致竹林成片枯死, 状如火烧^{[4, 5]}。现有研究表明^{[6, 7]}, 温度、湿度是刚竹毒蛾大发生的重要诱因, 冬季极端最低温的强度及持续时间同越冬代的发生程度存在密切关系, 若为暖冬, 低温持续时间短, 则越冬代发生严重, 反之则轻, 故温暖、湿润的环境有利于该虫害的发生, 而低海拔、 “ 倒春寒” 等区域则不利于其发生; 天敌寄生量与刚竹毒蛾虫口密度间存在显著正相关关系, 若虫口密度高、天敌少, 该虫则可能大暴发。以上研究对于刚竹毒蛾危害下的毛竹表征及虫害发生诱因的总结提供了重要参考, 但多局限于寄主外部形态与内部生理现象以及虫害发生机制的阐述及分析, 缺乏更深层次的总结。

光谱是光学遥感的基本信息, 光谱变化、不同健康状态下寄主光谱的可辨性是虫害遥感的理论基础, 从地面实测非成像高光谱数据出发, 挖掘虫害响应特征, 是“ 地-天” 耦合与虫害遥感监测、预警技术体系构建的必然途径。刚竹毒蛾危害下的毛竹叶片被食成缺刻, 导致竹体光合作用受阻, 其冠层光谱特征随之发生变化, 故总结不同虫害等级毛竹叶片的光谱变化机理是开展刚竹毒蛾危害检测研究的重要支点。高光谱数据波段多、波段宽度窄等特点为虫害解译提供了诸多信息, 但其存在数据量大、冗余度高等问题, 故挖掘对虫害响应敏感的光谱波长是开展寄主光谱变化机理研究的重要内容^{[8, 9, 10]}。

本文基于实测健康及刚竹毒蛾危害下的毛竹叶片高光谱数据, 通过分析不同虫害等级间的光谱特征差异, 确定特征波长, 以期为该虫害的遥感监测提供参考。

1 试验区概况

试验区位于福建省南平市顺昌县, 地理坐标为117° 29'— 118° 14'E, 26° 38'— 27° 121'N。该县土地总面积约1 985 km², 以山地、丘陵地貌为主; 常年受中亚热带海洋性季风气候影响, 年平均气温18.9 ℃, 无霜期约305 d, 雨日约164 d, 年均降水量1 600~1 800 mm, 日照1 740.7 h。区域内植被类型多, 森林资源丰富。截至2015年底, 该县植被覆盖率近95%, 森林覆盖率达79.75%, 其中毛竹林面积约3.21万hm², 占林地总面积的21.8%, 被誉为“ 中国竹子之乡” 。有关数据显示, 近10年来, 顺昌县各类林业有害生物年发生面积近3.33万hm², 防治面积3.13万hm²以上, 其中主要林业害虫包括马尾松毛虫、刚竹毒蛾、竹蝗等, 尽管已采取针对性措施控制虫害蔓延, 但时隔数年仍会发生大暴发, 而刚竹毒蛾等食叶害虫甚至出现年年大发生的趋势。

2 实验部分

2.1 光谱数据获取

于2016年6月25日— 7月5日在顺昌县内选择17个毛竹林小班, 利用合肥仪思特光电技术有限公司生产的ISI921VF-256野外地物光谱辐射计测定毛竹叶片光谱数据。该设备波长范围为380~1 050 nm, 计256个波段, 光谱分辨率为4 nm, 视场角为3° 。受测小班包括健康小班与刚竹毒蛾危害点, 测定时段处于刚竹毒蛾越冬代幼虫末期与第一代幼虫初期之间。

为保证光谱数据的准确性, 每当测量位置发生变化时进行一次标准白板校正; 每片竹叶分别测定近叶尖处、叶中、近叶基处3个部位, 每个部位取连续测定3次的平均值, 将3个部位的平均光谱数据作为该叶片光谱值, 并汇入数据库。刚竹毒蛾多产卵于竹叶背面或竹竿上, 幼虫出壳后噬食叶片, 致竹体水分缺失, 光合作用能力下降, 未被取食的竹叶及新生叶片则因养分缺失, 病斑严重, 故本文参照《植物病理学》^[11], 以病斑个数、大小及叶片缺刻程度为计量标准, 其与虫害等级对应关系为: (1)健康, 叶片无缺刻且无病斑; (2)轻度危害, 叶片无明显缺刻, 病斑小且少; (3)中度危害, 叶片有明显缺刻, 病斑小且多或大且少; (4)重度危害, 叶片缺刻严重, 病斑大且多。选择毛竹叶片高光谱数据105条, 并随机划分为两组, 其中71条作为实验组样本(健康19条、轻度危害16条、中度危害19条, 重度危害17条), 用于确定特征波长, 其余34条为验证组(健康9条、轻度危害8条、中度危害8条, 重度危害9条), 对结果予以检验。本文是刚竹毒蛾危害遥感响应机理研究的基础工作, 需要考虑与常用光学遥感影像的对接, 同时考虑光谱设备的设计, 保留900 nm以内的光谱数据。

2.2 光谱处理方法

利用ISI921VF-256野外地物光谱辐射计配套软件对光谱数据进行处理, 基于目标对象与白板光谱, 计算竹叶的光谱反射率。采用光谱微分法, 计算一阶微分及二阶微分光谱值。前人研究表明^{[12, 13]}, 经过微分处理的光谱数据能够有效放大光谱曲线中的细微变化, 进而提供更为丰富的光谱信息。一阶微分光谱可更好地表示原始光谱数据的变化率和极值点, 而二阶微分则突出了光谱曲线中切线斜率的变化速度。计算公式为

$\begin{array}{l} FD R_{λ_{i}} = \frac{d R}{d λ} = \frac{R_{λ_{i + 1}} - R_{λ_{i}}}{Δ λ} (1) \\ SD R_{λ_{i}} = \frac{d}{d λ} (\frac{d R}{d λ}) = \frac{R_{λ_{i + 2}} - 2 R_{λ_{i + 1}} + R_{λ_{i}}}{{(Δ λ)}^{2}} (2) \end{array}$

式中: FD $R_{λ_{i}}$ 为波段i和波段i+1之间的一阶微分光谱值; SD $R_{λ_{i}}$ 为波段i和波段i+1之间的二阶微分光谱值; λ _i为第i波段的波长值; $R_{λ_{i + 2}}$ , $R_{λ_{i + 1}}$ 与 $R_{λ_{i}}$ 分别代表λ _i₊₂, λ _i₊₁与λ _i波段处的原始光谱反射率; Δ λ 代表波段λ _i₊₁到λ _i之间的波长差值。

2.3 光谱特征响应波长确定方法

2.3.1 光谱数据分析及波长筛选规则

采用单因素方差分析法(One-way ANOVA)对光谱数据进行分析, 该方法是指不考虑个体间的差异, 仅涉及一个处理因素, 通过分析试验结果, 检验因素对数值型因变量是否存在显著性影响的方法。单因素方差分析法的核心在于计算SSA(组间离差平方和)和SSE(组内离差平方和), 其分别反映了控制变量和随机变量引起的离差; 随后计算总的变异平方和SST, 并进行F检验; 最后根据检验结果确定p值, 作出统计推断, p< 0.05表示不同水平下的控制变量其所属总体均值具有显著性差异, p< 0.01则表明差异达到极显著水平。 SSA, SSE, SST及F的计算公式为

$\begin{array}{l} SSA = \overset{k}{\sum_{i = 1}} n_{i} ({\overset{̅}{x}}_{i} {- \overset{̅}{x})}^{2} (3) \\ SSE = \overset{k}{\sum_{i = 1}} \overset{n_{i}}{\sum_{j = 1}} (x_{ij} - {\overset{̅}{x}}_{i})^{2} (4) \\ SST = SSA + SSE (5) \\ F = \frac{SSA / (k - 1)}{SSE / (n - k)} (6) \end{array}$

式中: k为水平数; n_i为第i个水平下的样本容量, 在此通过SPSS 22.0平台运行单因素方差分析。

结合研究需要, 建立筛选规则: (1)遥感影像易受大气散射、太阳光照条件变化、地形等多种因素的影响, 光谱信息易受干扰, 影响虫害识别, 筛选差异达到极显著水平(p< 0.01)的波长; (2)为尽量保证4个虫害等级各波长的可辨性, 对其(健康-轻度危害、健康-中度危害、健康-重度危害、轻度危害-中度危害、轻度危害-重度危害、中度危害-重度危害)进行两两分析, 若有4组(或以上)差异达极显著水平(p< 0.01), 则记录该波长; (3)较之ISI921VF-256光谱仪, 常用多光谱及高光谱遥感影像的光谱分辨率相对较低, 若所选波长范围过窄, 与遥感卫星波段的对应性较差, 考虑到“ 地-天” 不同尺度的耦合问题, 参考常用多光谱遥感卫星(Landsat, WorldView-2/3, RapidEye)及高光谱遥感卫星(EO-1 Hyperion)的波段设置, 对小于10 nm的波长予以剔除。

2.3.2 虫害判别能力分析方法

通过判断不同地物光谱间的相似度以确定其匹配程度已成为高光谱数据分类的主要方法^{[14, 15, 16]}。采用欧氏距离、相关系数及光谱角匹配等三种方法判定波长的虫害判别能力。

(1)欧式距离法: 是一种常用的相似性度量方法, 代表在m维空间中两点之间的真实距离(或者向量的自然长度), 通过二者之间的距离衡量其相似程度。计算公式为

$D_{ij} = \sqrt[]{\overset{n}{\sum_{i = 1, j = 1}} (X_{i} - X_{j})^{2}} (7)$

式中: D_ij为样本之间的欧式距离; X_i与X_j为第i和j个样本的光谱反射率(i, j=1, 2, …, n); n为样本数。

(2)相关系数法: 该系数可用以反映不同变量之间的相关密切程度, 系数越大表明二者的相似度越高。计算公式为

$r_{XY} = \frac{\overset{n}{\sum_{i = 1}} (X_{i} - \overset{̅}{X}) (Y_{i} - \overset{̅}{Y})}{\sqrt[]{\overset{n}{\sum_{i = 1}} (X_{i} {- \overset{̅}{X})}^{2} \overset{n}{\sum_{i = 1}} (Y_{i} {- \overset{̅}{Y})}^{2}}} (8)$

式中: r_XY为相关系数, X_i与Y_i为所取样本在第i波段处的光谱反射率(i=1, 2, 3, …, n), $\overset{̅}{X}$ 与 $\overset{̅}{Y}$ 为样本数据平均值; n为样本数。

(3)光谱角匹配法: 将光谱数据视为多维空间的向量, 据此计算试验光谱与参考光谱之间的向量夹角, 夹角越大(光谱角余弦值越低), 二者之间的相似度便越低, 反之亦然。计算公式为

$\cos α = \frac{XY}{| X || Y |} = \frac{\overset{n}{\sum_{i = 1}} (X_{i} Y_{i})}{\sqrt[]{\overset{n}{\sum_{i = 1}} X_{i}^{2}} \sqrt[]{\overset{n}{\sum_{i = 1}} Y_{i}^{2}}} (9)$

式中: α 为光谱向量夹角, 取值范围为[0, π /2]; cosα 为光谱角余弦值, 取值范围为[0, 1]; X_i与Y_i为向量在第i波段处的光谱反射率(i=1, 2, 3, …, n); n为样本数。

3 结果与讨论

3.1 毛竹叶片光谱单因素方差分析

3.1.1 原始光谱分析

基于实验组数据, 绘制各虫害等级叶片的光谱反射率曲线(图1)。可以看出, 健康叶片的光谱曲线在绿光波段处呈现一个反射小高峰, 至红光波段处出现反射率低谷, 而在可见光波段至近红外波段间光谱反射率陡增, 直至近红外波段处达到最高峰, 这与绿色植被理论光谱特征一致。

	Figure Option View Download New Window
	图1 各虫害等级毛竹叶片原始光谱反射率曲线Fig.1 Reflectance curves of moso bamboo leaves in different pest levels

受刚竹毒蛾危害后, 毛竹叶片光谱特征发生了较大变化。随着虫害等级的上升, 反射率不断降低, 具体表现为: (1)在可见光范围内, 光谱曲线中的“ 绿峰” 及“ 红谷” 是反映寄主健康程度的重要指标, 刚竹毒蛾危害下的毛竹叶片养分含量缺失, 其叶片光谱曲线中的“ 绿峰” 及“ 红谷” 有逐渐消失的趋势; (2)“ 红边” (680~780 nm)常作为虫害遥感的重要指示波段^{[17, 18]}, 可以看出, 受害叶片的“ 红边” 斜率明显减小, 且随着虫害等级的上升, 减小的趋势越发明显; (3)在近红外波段处, 植物的光谱特性主要受叶片内部构造影响, 健康植物叶片的叶肉海棉组织结构中存在许多空腔, 反射面较大, 故反射率较高, 当受害虫啃食时, 叶片细胞组织结构被破坏, 其反射率便随之降低, 虫害等级越高, 下降幅度也越大。利用单因素方差分析对原始光谱进行筛选, 获取差异达极显著水平(p< 0.01)的波长(图2), 可以看出, 各虫害等级叶片间的光谱差异较为显著, 其中具有极显著差异的波长多集中于近红外波段附近。

	Figure Option View Download New Window
	图2 各虫害等级毛竹叶片原始光谱单因素方差分析Fig.2 One-way ANOVA of original spectra of different pest levels

3.1.2 一阶微分光谱分析

对原始光谱进行一阶微分处理(图3)。可以看出: (1)各虫害等级间的光谱差异被放大, 具有极显著差异的波长数量大大增加; (2)总体而言, 受害叶片一阶微分光谱随着虫害等级的上升而降低, 表明受害叶片的光谱曲线在各波长位置上的斜率减小, 已逐渐失去了健康植被光谱的基本特征; (3)各虫害等级叶片的光谱曲线在“ 绿峰” 及“ 红谷” 处的变化率差异相对较大。同理获取一阶微分光谱中差异达极显著(p< 0.01)的波长(图4)。

	Figure Option View Download New Window
	图3 各虫害等级毛竹叶片一阶级分光谱反射率曲线Fig.3 Reflectance first derivative curves of moso bamboo leaves in different pest levels

	Figure Option View Download New Window
	图4 各虫害等级毛竹叶片一阶微分光谱单因素方差分析Fig.4 One-way ANOVA of first derivative spectra of different pest levels

3.1.3 二阶微分光谱分析

绘制二阶微分光谱反射率曲线(图5)。可以发现: (1)各虫害等级间的光谱差异进一步扩大, 差异达显著或极显著水平的波长数量有所增加; (2)由于二阶微分反映的是光谱曲线中切线斜率的变化速度, 故各虫害等级间具有极显著差异的波长范围较窄(均小于10 nm); (3)“ 红边” 及近红外波段处的差异相对更大。据此获取具有极显著差异(p< 0.01)的波长(图6)。

	Figure Option View Download New Window
	图5 各虫害等级毛竹叶片二阶微分光谱反射率曲线Fig.5 Reflectance second derivative curves of moso bamboo leaves in different pest levels

	Figure Option View Download New Window
	图6 各虫害等级毛竹叶片二阶微分光谱单因素方差分析Fig.6 One-way ANOVA of second derivative spectra of different pest levels

3.1.4 波长筛选

依据2.3.1建立的规则对以上波长进行筛选(表1), 并合并相邻波长, 可以看出, 经筛选后的波长主要涵盖绿光波段、 “ 红边” 及近红外波段。

表1 波长筛选结果 Table 1 Screening of characteristic wavelengths

3.2 虫害判别能力分析

3.2.1 基于欧式距离法的差异波长虫害判别能力分析

以健康叶片为准, 分别计算其与轻度危害、中度危害、重度危害叶片光谱间的欧式距离, 距离越大, 光谱差异也越大。绘制欧式距离散点图(图7), 可以看出, 随着虫害等级的上升, 健康叶片与受害叶片间所选波长的欧式距离均呈增大特征。为更清晰比较各虫害等级与健康等级间的差异, 计算欧式距离均值, 由表2可知, 在原始光谱的703.43~898.56 nm处, 健康-重度危害间距离最大, 健康-中度危害次之, 再次为健康-轻度危害, 而健康-健康的距离最小, 表明随着虫害等级的上升, 受害叶片和健康叶片在“ 红边” 和近红外波段处的反射率差异逐渐增大; 在一阶微分光谱的497.68~540.72, 554.53~585.25, 596.24~618.23及692.43~717.17 nm处所显示的规律与原始光谱一致, 表明在绿光波段及“ 红边” 位置上, 受害叶片和健康叶片间的光谱曲线变化率差异随虫害等级的上升而增大。

	Figure Option View Download New Window
	图7 欧氏距离法判别结果微点图Fig.7 Scatter plot of Euclidean distance between healthy and different pest levels (a): Original: 703.43~898.56 nm; (b): First derivative: 497.68~540.72 nm; (c): First derivative: 554.53~585.25 nm; (d): First derivative: 596.24~618.23 nm; (e): First derivative: 692.43~717.17 nm

表2 欧氏距离法判别结果 Table 2 Analysis of pest characteristic wavelengths with Euclidean distance method

3.2.2 基于相关系数法的差异波长虫害判别能力分析

基于相关系数法对所选波长进行虫害判别能力分析。由图8可知, 在原始光谱的703.43~898.56 nm处, 随着虫害等级的上升, 健康叶片与受害叶片的相关性呈现下降的趋势; 经一阶微分后, 各虫害等级叶片和健康叶片的相关性下降, 甚至出现了负相关, 表明受害毛竹叶片在不同波长处的反射率趋于紊乱, 其光谱特征发生了较大变化。经均值处理后发现(表3), 在原始光谱的703.43~898.56 nm处, 随着刚竹毒蛾危害程度的上升, 健康与各虫害等级间的相关系数减小; 经一阶微分后, 在497.68~540.72, 554.53~585.25及596.24~618.23 nm处, 健康-健康的相关系数最大, 其次为健康-轻度危害, 再次为健康-中度危害, 健康-重度最低, 表明受害叶片与健康叶片的光谱特征(切线斜率、曲线变化量等)在上述波长处的差异性随虫害等级的上升而增大。 692.43~717.17 nm处的判别结果出现异常, 表现为健康-健康相关系数最大, 健康-中度危害次之, 再次为健康-重度危害, 健康-轻度间相关性最小。

	Figure Option View Download New Window
	图8 相关系数法判别结果散点图Fig.8 Scatter plot of correlation coefficient in spectral sample with different healthy level (a): Original: 703.43~898.56 nm; (b): First derivative: 497.68~540.72 nm; (c): First derivative: 554.53~585.25 nm; (d): First derivative: 596.24~618.23 nm; (e): First derivative: 692.43~717.17 nm

3.2.3 基于光谱角匹配法的差异波长虫害判别能力分析

计算健康叶片和各虫害等级叶片的光谱角余弦值。由图9可知, 随着虫害等级的上升, 受害叶片与健康叶片在703.43~898.56 nm处的光谱角余弦值呈下降趋势; 经一阶微分后, 各虫害等级叶片在所选波长处的光谱差异被放大, 在596.24~618.23 nm处, 健康— 重度危害组中若干叶片间的光谱角余弦值达到负数(光谱角为钝角)。进行均值处理后, 得到与相关系数法类似的结果(表4), 在原始光谱的703.43~898.56 nm及一阶微分光谱的497.68~540.72, 554.53~585.25和596.24~618.23 nm处, 健康-重度危害间的光谱角余弦值最小, 即二者的光谱夹角最大, 相似性最小, 健康-中度危害次之, 再次为健康-轻度危害, 健康-健康间最小, 表明随着虫害等级的上升, 受害叶片与健康叶片间的光谱差异逐渐增大。在一阶微分光谱的692.43~717.17 nm处同样出现异常结果, 表现为健康-轻度危害的差异最大, 健康-重度危害次之, 再次为健康-中度危害, 健康-健康间的光谱角余弦值最小。

表3 相关系数法判别结果 Table 3 Analysis of pest characteristic wavelengths with correlation coefficient method

	Figure Option View Download New Window
	图9 光谱角匹配法判别结果散点图Fig.9 Scatter plot of spectral angle cosine value in spectral sample with different healthy level (a): Original: 703.43~898.56 nm; (b): First derivative: 497.68~540.72 nm; (c): First derivative: 554.53~585.25 nm; (d): First derivative: 596.24~618.23 nm; (e): First derivative: 692.43~717.17 nm

表4 光谱角匹配法判别结果 Table 4 Analysis of pest characteristic wavelengths with spectral angle mapping method

3.3 刚竹毒蛾危害下毛竹叶片特征波长的确定

综合上述分析可知, 各虫害等级叶片在原始光谱的703.43~898.56 nm及一阶微分光谱的497.68~540.72, 554.53~585.25和596.24~618.23 nm处差异显著, 且随着虫害等级的上升, 其与健康叶片的差异逐渐增大; 在一阶微分光谱692.43~717.17 nm处的特征与前者有所不同, 虽然其在欧式距离法的判别结果良好, 但相关系数、光谱角匹配法结果则显示, 健康叶片与受害叶片在该波长处的差异并未随虫害等级的上升而增大, 故予以剔除, 由此确定刚竹毒蛾危害下的毛竹叶片特征波长为: 原始光谱的703.43~898.56 nm及一阶微分光谱的497.68~540.72, 554.53~585.25和596.24~618.23 nm。

3.4 特征波长验证

利用验证组中的34条光谱数据, 分别用欧氏距离、相关系数及光谱角匹配法等三种方法对3.3中的结果予以验证。若所选特征波长通过2种(或以上)判别方法的检验, 则将其确定为特征波长。表5显示, 一阶微分光谱的三个特征波长(合并后)均通过了检验; 在原始光谱的703.43~898.56 nm位置上, 相关系数判别结果出现异常, 但其在欧式距离及光谱角匹配法的判别结果中表现良好。依据特征波长确定规则, 认为原始光谱的703.43~898.56 nm及一阶微分光谱的497.68~540.72, 554.53~585.25和596.24~618.23 nm均具备良好的虫害判别能力。

表5 特征波长验证结果 Table 5 Test of characteristic wavelengths

受母株年龄、土壤、地形、个体差异等因素影响, 即使是相同的虫害状态, 叶片的含水量、叶绿素、含氮量等各项理化参数亦各不相同, 由此导致光谱特征差异^{[19, 20, 21]}, 本文光谱数据亦显示, 同等级的毛竹叶片的光谱差异也可能较大。此外, 部分被刚竹毒蛾幼虫取食的叶片由于距离受害时间较近, 各养分指标尚未发生较大的变化, 其未被取食部分的光谱特征可能近似于健康叶片, 从研究结果中亦可看出, 若干受害叶片和健康叶片间的光谱差异较小。由此可见, 在划分虫害等级时, 若能实时获取其叶绿素含量、含氮量等营养指标, 通过设定相应权重进行综合考虑, 效果可能会更好。

遥感成像易受多种因素的干扰, 若选取的特征波长范围过窄, 则基于地面总结的规律可能在遥感影像中有所偏差, 考虑到“ 地-天” 不同尺度的对应效果, 结合各遥感卫星波段参数设置选取了4个特征波长, 为虫害遥感监测提供了理论支持。

4 结论

(1)随着虫害等级的上升, 毛竹叶片的光谱反射率随之下降, 整体上表现为: 健康> 轻度危害> 中度危害> 重度危害; (2)由于结构受损、养分缺失, 受害毛竹叶片的光谱特征发生变化, 其光谱曲线中的“ 绿峰” 和“ 红谷” 渐趋消失, “ 红边” 斜率下降, 且虫害等级越高, 该趋势越明显; (3)利用单因素方差分析获取各虫害等级间具有极显著差异的波长, 对其进行筛选后利用欧式距离、相关系数和光谱角匹配等三种方法分析各波长的虫害判别能力, 确定原始光谱的703.43~898.56 nm及一阶微分光谱的497.68~540.72, 554.53~585.25和596.24~618.23 nm为刚竹毒蛾危害下的毛竹叶片光谱特征波长, 其具备较强的虫害判别能力。

The authors have declared that no competing interests exist.

参考文献

文献选项

[1]	Radeloff V C, Mladenoff D J, Boyce M S. Remote Sensing of Envionment, 1999, 69(2): 156. [本文引用:1]
[2]	Latifi H, Schumann B, Kautz M, et al. Environmental Monitoring and Assessment, 2014, 186(1): 441. [本文引用:1]
[3]	ZHU Chen-hao, QU Shuai, ZHANG Xiao-li(朱程浩, 瞿帅, 张晓丽). Journal of Remote Sensing(遥感学报), 2016, 20(4): 653. [本文引用:1]
[4]	LIANG Guang-hong(梁光红). Journal of Fujian Forestry Science and Technology(福建林业科技), 2003, 30(1): 34. [本文引用:1]
[5]	CHEN De-liang, ZHAO Ren-you, QU Qiao-wen(陈德良, 赵仁友, 瞿巧文). Forest Pest and Disease(中国森林病虫), 2006, 25(6): 14. [本文引用:1]
[6]	WU Jian-qin(吴建勤). Plant Protection(植物保护), 2010, 36(3): 172. [本文引用:1]
[7]	LI Xing-tian(李兴天). Modern Horticulture(现代园艺), 2013, (1): 153. [本文引用:1]
[8]	Niemann K O, Quinn G, Stephen R, et al. Canadian Journal of Remote Sensing, 2015, 41(3): 191. [本文引用:1]
[9]	Ju Y W, Pan J, Wang X T, et al. Nematology, 2014, 16: 1197. [本文引用:1]
[10]	LIU Fei, FENG Lei, CHAI Rong-yao, et al(刘飞, 冯雷, 柴荣耀, 等). Acta Optica Sinica(光学学报), 2010, 30(2): 585. [本文引用:1]
[11]	Agrios G N. Plant Pathology(植物病理学). Beijing: China Agricultural Press(北京: 中国农业大学出版社), 2009. [本文引用:1]
[12]	LI Min-zan(李民赞). Spectral Analysis Techniques and Applications(光谱分析技术及其应用). Beijing: Science Press(北京: 科学出版社), 2006. [本文引用:1]
[13]	Jia J, Quan W. Ecological Informatics, 2016, 35: 1. [本文引用:1]
[14]	WU Wei-bin, ZHI Lei, HONG Tian-sheng, et al(吴伟斌, 支磊, 洪添胜, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2013, 29(10): 175. [本文引用:1]
[15]	WANG Qiang, YI Qiu-xiang, BAO An-ming, et al(王强, 易秋香, 包安明, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2012, 28(15): 125. [本文引用:1]
[16]	LI Ying, ZHANG Li-fu, YAN-Wei, et al(李颖, 张立福, 严薇, 等). Journal of Remote Sensing(遥感学报), 2013, 17(4): 855. [本文引用:1]
[17]	Pu R L, Gong P, Biging G S, et al. IEEE Transactions on Geoscience and Remote Sensing, 2003, 341(4): 916. [本文引用:1]
[18]	Cho M A, Debba P, Mutanga O, et al. International Journal of Applied Earth Observation and Geoinformation, 2012, 16(16): 85. [本文引用:1]
[19]	GONG Zhao-ning, ZHAO Ya-li, ZHAO Wen-ji, et al(宫兆宁, 赵雅莉, 赵文吉, 等). Acta Ecologica Sinica(生态学报), 2014, 34(20): 5736. [本文引用:1]
[20]	JI Rong-hua, ZHENG Li-hua, DENG Xiao-lei, et al(冀荣华, 郑立华, 邓小蕾, 等). Transactions of The Chinese Society of Agricultural Machinery(农业机械学报), 2014, 45(8): 269. [本文引用:1]
[21]	LIU Xiao-jun, TIAN Yong-chao, YAO Xia, et al(刘小军, 田永超, 姚霞, 等). Scientia Agricultura Sinica(中国农业科学), 2012, 45(3): 435. [本文引用:1]

1999

0.0

... 随着林业科技的进步尤其是林业遥感技术水平的不断提升, 以及生态文明战略的不断推进, 森林虫害遥感研究重新成为热点^[1,2,3] ...

2014

0.0

... 随着林业科技的进步尤其是林业遥感技术水平的不断提升, 以及生态文明战略的不断推进, 森林虫害遥感研究重新成为热点^[1,2,3] ...

2016

0.0

... 随着林业科技的进步尤其是林业遥感技术水平的不断提升, 以及生态文明战略的不断推进, 森林虫害遥感研究重新成为热点^[1,2,3] ...

2003

0.0

... 刚竹毒蛾是毛竹最主要的食叶害虫, 受害毛竹竹体瘦小, 竹节内积水, 产笋量急剧下降, 大发生时将竹叶食尽, 致竹林成片枯死, 状如火烧^[4,5] ...

2006

0.0

... 刚竹毒蛾是毛竹最主要的食叶害虫, 受害毛竹竹体瘦小, 竹节内积水, 产笋量急剧下降, 大发生时将竹叶食尽, 致竹林成片枯死, 状如火烧^[4,5] ...

2010

0.0

... 现有研究表明^[6,7], 温度、湿度是刚竹毒蛾大发生的重要诱因, 冬季极端最低温的强度及持续时间同越冬代的发生程度存在密切关系, 若为暖冬, 低温持续时间短, 则越冬代发生严重, 反之则轻, 故温暖、湿润的环境有利于该虫害的发生, 而低海拔、 #cod#x0201c ...

2013

0.0

2015

0.0

... 高光谱数据波段多、波段宽度窄等特点为虫害解译提供了诸多信息, 但其存在数据量大、冗余度高等问题, 故挖掘对虫害响应敏感的光谱波长是开展寄主光谱变化机理研究的重要内容^[8,9,10] ...

2014

0.0

2010

0.0

... 刚竹毒蛾多产卵于竹叶背面或竹竿上, 幼虫出壳后噬食叶片, 致竹体水分缺失, 光合作用能力下降, 未被取食的竹叶及新生叶片则因养分缺失, 病斑严重, 故本文参照《植物病理学》^[11], 以病斑个数、大小及叶片缺刻程度为计量标准, 其与虫害等级对应关系为: (1)健康, 叶片无缺刻且无病斑 ...

0.0

... 前人研究表明^[12,13], 经过微分处理的光谱数据能够有效放大光谱曲线中的细微变化, 进而提供更为丰富的光谱信息 ...

2016

0.0

... 前人研究表明^[12,13], 经过微分处理的光谱数据能够有效放大光谱曲线中的细微变化, 进而提供更为丰富的光谱信息 ...

2013

0.0

Wei-bin

, ZHI

Lei

, HONG

Tian-sheng

, et al(吴伟斌, 支磊, 洪添胜, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2013, 29(10): 175.

Abstract: Solenopsis invicta, wihich originated in Brazil, Argentina, Paraguay and the Panama Canal, has been spreading at the speed of 200 km every year since they invaded Mobile, Alabama in the United States from 1918 to 1931. Recently, Solenopsis invicta has indicated a tendency to rapidly spread all over the world, including China. They are polyphagous insects, but prefer high protein or high fat foods, such as the larvae of some small insects or vertebrates and seeds or fruits with high oil. Therefore, they are severely dangerous to human health, public security, agriculture, forestry production, and the ecological environment in the invaded areas.Solenopsis invicta is social insect of the family Formicidae, Hymenoptera with the habit of gregarious nesting. They are strongly offensive against intruders, and they will attack the intruders in full strength once their nest is disturbed. Thus, some experts think that the threat of Solenopsis invicta in China cannot be ignored and that corresponding measures should be imminently taken to prevent the Solenopsis invicta invasion. Research on the detection method of a Solenopsis invicta nest to eliminate Solenopsis invicta from the source is of practical significance to the prevention of a Solenopsis invicta disaster. Therefore, finding out the location of a nest of Solenopsis invicta is of more practical value than only exterminating Solenopsis invicta.The feasibility of detection of a nest of Solenopsis invicta based on spectrum analysis technology was examined by using an optical spectrum instrument with reflect characteristics. A differential coefficient method and a logarithmic method to analyze original spectral reflectivity were used. The method of using character factors was studied to get the best waveband for distinguishing the ant nest from other things. An Eculidean distance method was used to calculate the average distance, and effective band of the nest of Solenopsis invicta's recognized effect. In addition, the average of the Euclidean distance of different features is greater than the average of the Euclidean distance of the same feature, and the samples, which were collected and measured by the laboratory hyper-spectral imager to verify the reliability of the selected band. Finally, the results were compared with the traditional method. We randomly selected eight sample points in the Wushan Square of South China Agricultural University. Using the traditional method and a spectrum analytical method to prove it, we got the same result. However, it is more convenient and faster to get the information by using the spectrum analytical method.As a result, the effective wavebands to recognize the Solenopsis invicta nest mostly concentrate on 701-1 510 nm, which are mainly red light bands and near-infrared bands. Additionally, the observation shows that the original spectrum and the logarithmic method could be used to distinguish the nest of Solenopsis invicta from common soil and grass. Moreover, the latter can make it clearer than the others, while the spectrum first-order differential method and the logarithmic after a differential transformation could distinguish between grass and ant nest soil, or grass and common soil, but not between Solenopsis invicta nest soil and common soil. This observation may provide useful information for detecting the Solenopsis invicta nest.

为探讨利用光谱分析技术对红火蚁蚁巢进行探测识别的可行性，该文利用高光谱仪采集红火蚁蚁巢光谱特征参数，采用微分法、对数法对原始光谱反射率进行处理，通过提取光谱特征参数的方法得到识别红火蚁蚁巢的最佳波段，确定识别不同地物的波段大多位于红光波段和近红外波段，分布在701～1 510 nm之间。利用欧氏距离法对识别效果进行验证，并通过实验室高光谱成像仪进行室内参照试验，对有效波段的可靠性加以验证。结果表明，原始光谱以及光谱对数变换方法能够有效区分红火蚁蚁巢土及周边不同地物，而一阶微分和对数后进行一阶微分变换得到的欧式距离可以区分草和蚁巢土及草和普通土，但不能区分蚁巢土与普通土。研究结果可为红火蚁蚁巢的快速检测提供参考。

... 通过判断不同地物光谱间的相似度以确定其匹配程度已成为高光谱数据分类的主要方法^[14,15,16] ...

2012

0.0

WANG

Qiang

, YI

Qiu-xiang

, BAO

An-ming

, et al(王强, 易秋香, 包安明, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2012, 28(15): 125.

In order to further improve the estimation accuracy of cotton chlorophyll density by hyperspectral reflectance, canopy hyperspectral reflectance and chlorophyll density were recorded at four different growth stages of cotton in a field experiment. All two-band combinations (350 to 1100 nm) in the ratio type of vegetation index (RVI) and the normalized difference type of vegetation index (NDVI) were performed on raw spectral reflectance and the first derivative reflectance, and then the correlation between all two-band combinations and cholorophyll density were determined. The coefficients (r) were presented in matrix plots. Basing on the results of correlation analysis, the estimation models of chlorophyll density were established using linear regression and multiply stepwise regression methods, and then the predictive power of four predictors were analyzed, i.e. single narrow band raw reflectance and the first derivative reflectance, the established vegetation indices for chlorophyll density estimation, and the optimal band combination vegetation indices. Three main conclusions were obtained: 1) The performance of first derivative reflectance was evidently better than raw reflectance; 2) The precision and stability of estimation models based on vegetation indices were normally much higher than models based on single band or multiply bands; 3) Among four types independent variables, DR756 was the best candidate for single-band models, ratio index DR635/DR643 and normalized difference index (DR1055-DR684)/(DR1055+DR684) were the best among all band combination indices. In conclusion, the model based on DR635/DR643 obtained the most satisfied results for the estimation of chlorophyll density, and the correlation coefficient between estimated and measured chlorophyll density reached 0.821. The study will provide a reference for the better application of hyperspectral reflectance in chlorophyll density derivation.

为了进一步提高棉花叶绿素密度高光谱估算精度，该研究以棉花冠层叶绿素密度以及冠层高光谱反射率为数据源，在分析叶绿素密度与原始高光谱反射率（R）、一阶导数光谱反射率（DR）、已有光谱指数及全波段组合指数相关性的基础上，采用线性及多元逐步回归技术构建了叶绿素密度高光谱诊断模型，系统对比分析了以上4种光谱形式用于棉花冠层叶绿素密度诊断的精度。结果表明：1）基于一阶导数光谱反射率的估算模型精度明显优于原始光谱反射率；2）基于比值指数或归一化指数形式的估算模型精度及稳定性要优于单波段或多波段的线性模型；3）单波段变量DR756、全波度组合比值指数DR635/DR643以及归一化指数(DR1055-DR684)/(DR1055+DR684)均可较好的实现叶绿素密度估算，其中由DR635/DR643为自变量的模型所得到棉花冠层叶绿素密度估算值与实测值拟合最好，相关系数达到0.821。该研究可为高光谱技术在棉花冠层叶绿素密度诊断中的更好应用提供参考。

... 通过判断不同地物光谱间的相似度以确定其匹配程度已成为高光谱数据分类的主要方法^[14,15,16] ...

2013

0.0

Ying

, ZHANG

Li-fu

, YAN-Wei, et al(李颖, 张立福, 严薇, 等). Journal of Remote Sensing(遥感学报), 2013, 17(4): 855.

Weed identification is a basic task in precision agriculture, as well as in the principle of variable spraying and accurate weeding. Field imaging spectrometer data, with both hyperspectral and high spatial resolutions, have potential applications in weed identification. Currently, methods for weed identification include considering shape features based on machine vision, which perform poorly when weeds and crops have similar shape features, while the use of hyperspectral features have low efficiency when it is identified in pixels. To overcome the limitations of these existing methods, a new weed identification method combining the strengths of both object-oriented and spectral feature matching approaches is proposed. The proposed method extracts and analyzes the shape features and the spectral curves of plant object samples, and builds a decision tree using shape feature rules and spectral angle mapper to identify the plant objects in the experimental field. The results show that the proposed method could identify different kinds of plant objects with similar shape features by using hyperspectral features, and could overcome the difficulties in identifying same objects with different spectra and different objects with the same spectrum by using shape feature rules. The identification accuracy of the described method is higher than both the spectral angle mapper method and the color and shape analysis.

地面成像光谱数据兼具高光谱分辨率与高空间分辨率,在田间杂草识别中具有很好的应用前景。目前基于机器视觉的杂草识别方法以形状特征为主,当作物杂草形态相似时识别的困难和利用高光谱特征以像元为单元识别时效率较低,不利于实时自动化除草,因此,本文提出一种综合面向对象与高光谱特征匹配的杂草识别方法,在对作物杂草对象样本的形状特征和光谱曲线提取分析的基础上,建立基于形状特征规则与光谱角匹配的植物对象识别决策树,用于识别实验田中的作物杂草对象。实验结果表明,当场景中某些不同种类植物对象的形态相似时,基于形状特征规则与光谱角匹配的杂草识别方法可借助高光谱特征精细区分植物对象的种类,且在形状特征规则约束下使用高光谱特征匹配法识别植物对象,可克服"同物异谱"和"同谱异物"现象带来的不确定性,该方法识别精度可优于仅使用光谱角匹配法的情况,并优于使用颜色和形状分析技术的情况。

... 通过判断不同地物光谱间的相似度以确定其匹配程度已成为高光谱数据分类的主要方法^[14,15,16] ...

2003

0.0

... (680~780 nm)常作为虫害遥感的重要指示波段^[17,18], 可以看出, 受害叶片的#cod#x0201c ...

2012

0.0

... (680~780 nm)常作为虫害遥感的重要指示波段^[17,18], 可以看出, 受害叶片的#cod#x0201c ...

2014

0.0

GONG

Zhao-ning

, ZHAO

Ya-li

, ZHAO

Wen-ji

, et al(宫兆宁, 赵雅莉, 赵文吉, 等). Acta Ecologica Sinica(生态学报), 2014, 34(20): 5736.

Chlorophyll can be an indicator in photosynthesis capacity and vegetation developmental stages,which is also one of important indicators to monitor health status of wetland vegetation growth. Hyperspectral remote sensing technology can provide a simple, effective and non-destructive data acquisition, which can offer processing method for quantifying diagnosis plant chlorophyll content as well. This study used the Fieldspec 3 spectrometer and a plant probe leaf clip spectral detector to guarantee the spectrum are detected in the same area of the leaf, it is also eliminating the background reflectance, spectral fluctuations caused by bending of the blade surface and the impact caused by leaf internal variability. This study determined the typical wetland plants leaf hyperspectral reflectance data at Wild Duck Lake, and at the same time the corresponding leaf chlorophyll content was measured using a spectrophotometer indoor. The relationship between chlorophyll content and the Trilateral parameters, as well as the ratio of spectral index model (SR) and normalized difference spectral index (ND) were established respectively using linear regression model., then 3-Fold Cross Validation(3K-CV) was used to test the accuracy of the estimation model. The results showed that most of the "trilateral" parameters were significantly correlated with plant leaf chlorophyll content; the maximum correlation coefficient reached 0.867. The correlation coefficient between ratio (SR) and normalized (ND) and chlorophyll content were high in general. Suitable band combinations were 550-700 nm,700-1400nm, 550-700 nm and 1600-1900 nm. The best indices with highest correlation with chlorophyll content were SR (calculated from bands 565 nm and 740 nm) and ND (calculated from bands 565 nm and 735 nm). And then by choosing the best correlation spectrum characteristic parameters based on the Trilateral parameters and ND model index, a plant chlorophyll estimation model was constructed. Among them,a chlorophyll content estimation model established by Red edge position (WP_r) of spectral characteristic parameters and ND (565nm, 735nm) spectral index achieved better test results, and R 2 both reached above 0.8, the estimation model were y =0.113 x -78.74, y =5.5762 x +4.4828. Using 3K-CV method for testing and validation, the prediction accuracies of both plant leaf chlorophyll content estimation models were 93.9% and 90.7%, respectively. The quantitative analysis of hyperspectral remote sensing technology shows a strong advantage in detecting vegetation weak spectral differences and provides an important theoretical basis and technical support for the practical application in the diagnosis of plant chlorophyll content.

叶绿素是光合作用能力和植被发育阶段的指示器，是监测湿地植被生长健康状况的重要指标之一；高光谱遥感技术可以为植物叶绿素含量的定量化诊断提供简便有效、非破坏性的数据采集和处理方法。为保证被探测叶片面积相同，消除背景反射、叶片表面弯曲造成的光谱波动及叶片内部变异造成的影响，研究采用Field Spec 3光谱仪加载手持叶夹式叶片光谱探测器，测定野鸭湖湿地典型植物的叶片高光谱反射率数据，同时通过分光光度计室内测定相应叶片的叶绿素含量。采用相关性及单变量线性拟合分析技术，建立二者的关系模型，包括叶绿素含量与"三边"参数的相关模型以及比值光谱指数（SR）模型和归一化差值光谱指数（ND）模型，并采用交叉检验中的3K-CV方法对估算模型进行模型精度检验。结果表明：植物叶片叶绿素含量与"三边"参数大多都呈极显著相关，相关系数最大达到0.867；计算光谱反射率组成的比值（SR）和归一化（ND）光谱指数与叶绿素含量的决定系数，总体相关性比较高，较好的波段组合均为550-700nm与700-1400nm以及550-700nm与1600-1900nm，与叶绿素含量相关性最好的指数分别是SR（565nm，740nm）和ND（565nm，735nm）。并通过选取相关性最佳的光谱特征参数，分别基于"三边"参数和ND模型指数构建了植物叶片叶绿素含量的估算模型。其中，基于红边位置（WP_r）光谱特征参数和ND（565nm，735nm）光谱指数建立的叶绿素含量估算模型，取得了较好的测试效果，检验拟合方程的决定系数（ R 2 ）都达到0.8以上，估算模型分别为 y =0.113 x -78.74， y =5.5762 x +4.4828。通过3K-CV方法进行测试和检验，植物叶绿素含量估算模型均取得了较为理想的预测精度，预测精度的分别为93.9%及90.7%。高光谱遥感技术对植被进行微弱光谱差异的定量分析，在植被遥感研究与应用中表现出强大优势，为植物叶绿素含量诊断中的实际应用提供了重要的理论依据和技术支持。

... 受母株年龄、土壤、地形、个体差异等因素影响, 即使是相同的虫害状态, 叶片的含水量、叶绿素、含氮量等各项理化参数亦各不相同, 由此导致光谱特征差异^[19,20,21], 本文光谱数据亦显示, 同等级的毛竹叶片的光谱差异也可能较大 ...

2014

0.0

2012

0.0

LIU

Xiao-jun

, TIAN

Yong-chao

, YAO

Xia

, et al(刘小军, 田永超, 姚霞, 等). Scientia Agricultura Sinica(中国农业科学), 2012, 45(3): 435.

【Objective】 The objective of the experiments is to develop a key method for fast and nondestructive monitoring leaf water content (LWC) in rice (Oryza sativa L.). 【Method】 Two field experiments were conducted with different soil water conditions and rice cultivars across two growing seasons, and time-course measurements were taken on leaf hyperspectral reflectance and LWC at top four leaves over main growth stages. Several kinds of hyperspectral indices at leaf scale including ratio spectral indices (RSI), normalized difference spectral indices (NDSI) and difference spectral indices (DSI) with all combinations of two wavebands between 350 and 2 500 nm were calculated, and their relationships to LWC were analyzed. 【Result】 The results indicated that the leaf spectral reflectance varied distinctly with soil water treatments and different top leaves, the sensitivity bands mostly occured within near-infrared and short-infrared spectral regions. The spectral indices as RSI (R1402, R2272) and NDSI (R1402, R2272) were linear with LWC, giving the determination coefficient of linear regression (S-R2) of 0.80, and the predicted R2 (P-R2) based on the testing performance with independent datasets as 0.86. 【Conclusion】 It is concluded that the RSI (R1402, R2272) and NDSI (R1402, R2272) can be used to monitor leaf water content in rice.

1.南京农业大学农学院/国家信息农业工程技术中心/江苏省信息农业高技术研究重点实验室，南京210095

【目的】建立快速、无损诊断水稻叶片含水量的估测模型，为水稻水分精确管理提供依据。【方法】基于2年不同土壤水分处理和水稻品种的池栽试验，于水稻主要生育时期同步测定顶部4张叶片的光谱反射率和含水量，系统分析350—2 500 nm波段范围内任意两波段组合而成的比值（RSI）、归一化差值（NDSI）及差值（DSI）光谱指数，并分析其与叶片含水量的量化关系。【结果】不同土壤水分处理和叶位间，叶片反射光谱具有显著的时空变化特征，叶片含水量的敏感光谱波段主要位于近红外及短波红外区域；RSI （R1402, R2272）及NDSI （R1402, R2272）光谱指数与叶片含水量呈现良好的线性相关，线性拟合R2均达到0.80。基于独立试验资料对所建模型进行测试检验也显示，预测值和观察值的拟合R2也均达到0.86。【结论】RSI（R1402, R2272）、NDSI（R1402, R2272）均可用于水稻叶片含水量的定量监测。