基于极限学习机的浅海水深遥感反演研究  

Research on Remote Sensing Inversion of Shallow Water Depth Based on Extreme Learning Machine

作　　者：吴忠强 毛志华 王正[1,2,3] 邱耀炜 沈蔚 WU Zhongqiang;MAO Zhihua;WANG Zheng;QIU Yaowei;SHEN Wei(School of Geographic and Oceanographic Sciences,Nanjing University,Nanjing 210023,China;Collaborative Innovation Center for the South China Sea Studies,Nanjing University,Nanjing 210023,China;States Key Laboratory of Satellite Ocean Environment Dynamics,Second Institute of Oceanography,Ministry of Natural Resources,Hangzhou 310012,China;College of Marine Sciences,Shanghai Ocean University,Shanghai 201306,China;Estuarine and Oceanographic Mapping Engineering Research Center of Shanghai Municipal Ocean Bureau,Shanghai 201306,China)

机构地区：[1]南京大学地理与海洋科学学院,江苏南京210023 [2]中国南海研究协同创新中心,江苏南京210093 [3]自然资源部第二海洋研究所卫星海洋环境动力学国家重点实验室,浙江杭州310012 [4]上海海洋大学海洋科学学院,上海201306 [5]上海市海洋局河口海洋测绘工程技术研究中心,上海201306

出　　处：《海洋测绘》2019年第3期11-15,共5页Hydrographic Surveying and Charting

基　　金：国家重点研发计划项目(2016YFC140091);国家海洋公益项目(201005030);国家自然科学基金(41476156;41621064);高分辨率对地观测系统重大专项(41-Y20A31-9003-15/17)

摘　　要：以Sentinel-2A遥感影像为数据源,在数据预处理基础上,使用极限学习机建立水深反演非线性回归模型,并与双波段回归模型、神经网络BP模型进行了比较。结果表明,总体上,支持极限学习机反演精度比较高,在水深5~20m处,极限学习机模型具有较高的反演精度和良好的效果,适用于研究区的水深反演。Taking Sentinel-2A image as data sources,on the basis of data preprocessing,the depth inversion nonlinear regression model is established using Extreme Learning Machine(ELM),and a comparison is made between dual-band regression model and BP neural network model respectively.The results show that,in general,the inversion accuracy of support vector machine(SVM)is higher,and in depth of 5-10-15-20,ELM has a higher inversion accuracy and better effect,which suits depth inversion in research ares.

关 键 词：多光谱遥感 Sentinel-2A 水深反演 双波段比值法 神经网络BP 极限学习机 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]