| 基于神经网络方法的高光谱遥感浅海水深反演 |
| |
| 引用本文： | 施英妮,张亭禄,周晓中,吴耀平,石立坚. 基于神经网络方法的高光谱遥感浅海水深反演[J]. 高技术通讯, 2008, 18(1): 71-76. DOI: 10.3772/j.issn.1002-0470.2008.01.014 |
| |
| 作者姓名： | 施英妮 张亭禄 周晓中 吴耀平 石立坚 |
| |
| 作者单位： | 总参作战部,0526,工程办公室,北京,100081;中国海洋大学海洋遥感研究所,青岛,266003;中国海洋大学海洋遥感研究所,青岛,266003;总参作战部,0526,工程办公室,北京,100081 |
| |
| 摘 要： | 利用人工神经网络方法进行了高光谱遥感反演浅海水深的初步研究.在产生模拟数据时,为保证模拟数据的合理性,引入了根据水体和海底特性来划分光学浅水和光学深水的方法,并初步研究了利用光谱徽分技术进行光学浅水和光学深水区分的有效性.在人工神经网络建模过程中,采用主成分分析的方法对网络的输入数据进行预处理,显著提高了网络的学习速度.建立的人工神经网络模型和基于非线性最优化方法的反演算法与实测数据的反演结果相比较,人工神经网络模型的反演精度明显高于非线性最优化反演算法.
|
| 关 键 词： | 浅海水深 高光谱遥感 人工神经网络 光谱微分技术 主成分分析 |
| 修稿时间： | 2006-11-14 |
Study on hyperspectral remote sensing for mapping shallow waters depth with neural network method |
| |
| Shi Yingni,Zhang Tinglu,Zhou Xiaozhong,Wu Yaoping,Shi Lijian. Study on hyperspectral remote sensing for mapping shallow waters depth with neural network method[J]. High Technology Letters, 2008, 18(1): 71-76. DOI: 10.3772/j.issn.1002-0470.2008.01.014 |
| |
| Authors: | Shi Yingni Zhang Tinglu Zhou Xiaozhong Wu Yaoping Shi Lijian |
| |
| Affiliation: | Shi Yingni~(* **) Zhang Tinglu~(**) Zhou Xiaozhong~* Wu Yaoping~* Shi Lijian~(**)(* Engineering Office 0526,Department of General Staff,Beijing 100081)(** Ocean Remote Sensing Institute of Ocean University of China,Qingdao 266003) |
| |
| Abstract: | The research about hyperspeetral remote sensing for shallow water depth based on artificial neural hetwork(ANN) technique was carried out.Firstly the method to distinguish optical shallow waters from optical deep waters based on the inherent optical properties of waters and bottom features was introduced for insuring the rationality of the simulated data. Then the validity of the method to differentiate between the two kinds of waters with the spectral derivative technique was studied.In the establishment o... |
| |
| Keywords: | shallow waters depth hyperspectral remote sensing artificial neural network spectral derivative principal component analysis |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
