| 引用本文: | 况润元,曾帅,赵哲,肖阳.基于实测高光谱数据的鄱阳湖湿地植被光谱差异波段提取.湖泊科学,2017,29(6):1485-1490. DOI:10.18307/2017.0620 |
| KUANG Runyuan,ZENG Shuai,ZHAO Zhe,XIAO Yang.Extraction of the discriminative bands of Lake Poyang wetland vegetation based on the measured hyperspectral data. J. Lake Sci.2017,29(6):1485-1490. DOI:10.18307/2017.0620 |
|
| 摘要: |
| 高光谱遥感技术的出现为有效解决湿地植被种类的精细识别和分类提供了可能.通过实地测取鄱阳湖湿地5种植被的高光谱数据，在对数据预处理的基础上，提出一种基于数据误差范围和植被光谱均值差的植被光谱差异波段提取方法.将该方法应用于包络线变换前后的光谱曲线提取植被的光谱差异波段，最后利用马氏距离法检验植被识别效果.结果表明：本文中的方法有效提取了植被光谱差异波段，其中变换前光谱差异波段分别为663~688 nm，变换后为581~636、660~695和1225~1236 nm.在光谱差异波段范围内，同种植被的马氏距离值小于异种植被的马氏距离值，可有效对植被进行识别.研究结果为湿地植被分类识别奠定了理论基础，同时为湖泊湿地植被以及湖泊生态环境的保护决策提供科学依据. |
| 关键词: 鄱阳湖湿地 植被光谱 光谱差异波段 |
| DOI：10.18307/2017.0620 |
| 分类号: |
| 基金项目:国家自然科学基金项目（41101322）、江西省自然科学基金项目（20114BAB213022）和江西省教育厅科学技术研究项目（GJJ160617）联合资助. |
|
| Extraction of the discriminative bands of Lake Poyang wetland vegetation based on the measured hyperspectral data |
|
KUANG Runyuan, ZENG Shuai, ZHAO Zhe, XIAO Yang
|
|
School of Architectural and Surveying & Mapping Engineering, Jiangxi University of Science and Technology, Ganzhou 341000, P. R. China
|
| Abstract: |
| Identification and classification of wetland vegetation by spectral data is often one of the key points and difficulties in vegetation remote sensing. Compared with multispectral data, hyperspectral ones have the characteristics of more bands, larger amount of data and higher redundancy. The emergence of hyperspectral remote sensing technology provides chance for solving problems of identifying and classifying wetland vegetation species precisely. For the discrimination of wetland vegetation species with hyperspectral remote sensing technology, the extraction of the discriminative bands is very important. The hyperspectral data on Acorus calamus, Cynodon dactylon, Carex cinerascens, Persicaria hydropiper, Phalaris arundinacea in Lake Poyang wetland are measured by using ASD FieldSpec 4 spectrometers. A new method is proposed to extract the discriminative bands between vegetation species of original spectra and envelope spectra. Finally, we use the method of Mahalanobis distance to examine the vegetation identification effects. The research results show that:The new method can extract the discriminative bands between vegetation species effectively. The discriminative bands of original spectra is 663-688 nm, and the discriminative bands of envelope spectra are:418-451 nm, 487-507 nm, 621-637 nm, 658-85 nm, 1222-1245 nm. In those discriminative bands, the Mahalanobis distance values between the same species is less than those among different species, so it could help us to identify different types of vegetation species effectively. The research results lay a foundation for the wetland vegetation classification, and provide a scientific basis for the protection of the lake wetland vegetation and the ecological environment of lakes. |
| Key words: Lake Poyang wetland hyperspectral data discriminative bands |
