| 基于无人机高光谱特征的红树林种群识别研究 |
| |
| 引用本文： | 周在明,陈本清,徐冉,方维.基于无人机高光谱特征的红树林种群识别研究[J].海洋学报,2021,43(9):137-145. |
| |
| 作者姓名： | 周在明 陈本清 徐冉 方维 |
| |
| 作者单位： | 1.自然资源部第三海洋研究所 海洋声学与遥感实验室，福建 厦门 361005 |
| |
| 基金项目： | NSFC-山东联合基金（U1806203） |
| |
| 摘 要： | 红树林种群的组成和分布对于红树林生态系统的保护和恢复至关重要。本研究以漳江口红树林保护区为研究对象，通过获取无人机高光谱影像，进行光谱特征分析、光谱微分变换和包络线去除，提取了911组17个光谱特征参数，通过逐步判别分析筛选出13个用于决策树构建的特征参数，最终通过C5.0决策树模型获得了研究区红树林种群的分布状况。结果表明，漳江口红树林保护区植被种群呈现自上到下不同类型的分布情况，研究区上部以桐花树和秋茄混合类型为主，中间区域呈现白骨壤、桐花树和秋茄三者共生的现状，研究区下部则以白骨壤分布为主，伴生有少量的秋茄。通过混淆矩阵计算，得到研究区总体分类精度为 87.95%，Kappa系数为 83.81%，具有较好的精度。研究结果可为区域红树林湿地保护提供数据支撑，为红树林种群识别研究提供方法参考。
|
| 关 键 词： | 红树林 漳江口 无人机 高光谱影像 种群识别 |
| 收稿时间： | 2020-12-08 |
Identification of the mangrove species using UAV hyperspectral images: A case study of Zhangjiangkou mangrove national nature reserve |
| |
| Affiliation: | 1.Ocean Acoustics and Remote Sensing Laboratory, Third Institute of Oceanography, Ministry of Natural Resources, Xiamen 361005, China2.Forestry College, Fujian Agriculture and Forestry University, Fuzhou 350002, China3.Zhangjiangkou Mangrove National Nature Reserve, Yunxiao 363000, China |
| |
| Abstract: | The composition and distribution of mangrove species are crucial to the protection and restoration of mangrove wetland ecosystems. In this study, mangrove species distribution was identified by unmanned aerial vehicle (UAV) hyperspectral images from Zhangjiangkou mangrove national nature reserve. Spectral characteristics, spectral differential, and spectral continuum removal were analyzed, 17 spectral parameters of 911 group spectral data from different vegetation species were obtained. Furthermore, 13 parameters for decision tree construction were selected by stepwise discriminant analysis. As a result, an accurate distribution map of mangrove species in the study area was obtained through C5.0 decision tree classification model. The vegetation species present different distribution types from top to bottom in the Zhangjiangkou mangrove national nature reserve. The upper part of the study area was dominated by the mixed type of Aegiceras corniculatum and Kandelia obovata. The middle area showed symbiosis status of three different mangrove species Avicennia marina, Aegiceras corniculatum and Kandelia obovata. The lower part of the study area was dominated by Avicennia marina, and a small amount of Kandelia obovata. Through the confusion matrix, the overall classification accuracy is 87.95% and the Kappa coefficient is 83.81%, showed a satisfactory precision. Therefore, our mangrove species identification results from UAV hyperspectral images could be used as a reference for ecological protection of regional mangrove wetland, and also as a identification method reference for mangrove species. |
| |
| Keywords: | |
|
| 点击此处可从《海洋学报》浏览原始摘要信息 |
|
点击此处可从《海洋学报》下载全文 |
|
