基于FOD和最优光谱特征的高光谱水质参数反演模型  

Inversion Model of Hyperspectral Water Quality Parameters Based on FOD and Optimal Spectral Characteristics

作　　者：张雨晴 赵起超 刘其悦 房红记 韩文龙 陈雯玥 ZHANG Yu-qing;ZHAO Qi-chao;LIU Qi-yue;FANG Hong-ji;HAN Wen-long;CHEN Wen-yue(North China Institute of Aerospace Engineering Institute of Remote Sensing Information Engineering,Langfang 065000,China;Hebei Aerospace Remote Sensing Information Processing and Application Collaborative Innovation Center,Langfang 065000,China;Langfang Vilda Software Co.,Ltd.,Langfang 065000,China;Zhongke Space Information(Langfang)Research Institute,Langfang 065000,China)

机构地区：[1]北华航天工业学院遥感信息工程学院,河北廊坊065000 [2]河北省航天遥感信息处理与应用协同创新中心,河北廊坊065000 [3]廊坊维尔达软件股份有限公司,河北廊坊065000 [4]中科空间信息(廊坊)研究院,河北廊坊065000

出　　处：《光谱学与光谱分析》2025年第3期842-851,共10页Spectroscopy and Spectral Analysis

基　　金：高分辨率对地观测系统国家科技重大专项(67-Y50G04-9001-22/23,67-Y50G05-9001-22/23);河北省教育厅科学技术研究项目(CXY2023011);北华航天工业学院研究生院创新资助项目(YKY-2023-71)资助。

摘　　要：准确高效获取水体叶绿素a(Chl-a)浓度是水体富营养化情况改善和可持续发展的前提。以水面野外高光谱反射率和实测水体Chl-a浓度为数据源,利用分数阶微分(FOD)对原始光谱反射率进行步长为0.1的处理,通过探寻光谱数据与水体Chl-a浓度之间的相关性来筛选特征波段,构建可变阶光谱数据集。使用偏最小二乘模型(PLS)筛选最优特征构建数据集;按照7∶3的比例划分为建模集和验证集;采用支持向量机(SVM)和深度森林(DF)模型建立水体Chl-a浓度反演模型,并与采用原始数据直接构建模型以及常见降维方法构建的模型进行对比验证。结果显示:FOD技术可以在一定程度上减弱高光谱噪声并挖掘潜在光谱信息,提高高光谱反射率与水体Chl-a浓度的相关性。利用FOD结合PLS先筛选特征再建立的水体Chl-a浓度反演模型相较于利用原始数据以及PCA降维后建立的水体Chl-a反演模型来说,R^(2)均有所提高,均方误差MSE和平均绝对误差MAE均有所下降。其中DF相对于其他三种模型具有较高的拟合度,预测精度也相对较高,建模集R^(2)=0.96,MSE=0.51μg·L^(-1),MAE=0.64μg·L^(-1);验证集R^(2)=0.89,MSE=0.69μg·L^(-1),MAE=0.64μg·L^(-1)。总体来看,基于FOD重组后的可变阶光谱数据集和PLS优选特征建立水体Chl-a浓度反演模型是可行的;对比分析构建的其他模型反演结果发现,DF对水体Chl-a的反演效果较好。该工作为内陆二类水体Chl-a高光谱遥感反演提供一定的理论依据和技术支持,助力白洋淀水质持续监测,也为以后高光谱卫星遥感影像反演Chl-a提供新思路。Accurate and efficient acquisition of chlorophyll-a(Chl-a)concentration in water is the prerequisite for improving eutrophication and sustainable development of water bodies.This study used the hyperspectral reflectance of the water surface and the measured Chl-a water concentration as data sources.The original spectral reflectance was processed with a step size 0.1 by fractional order differentiation(FOD)technology.The characteristic bands were screened by exploring the correlation between the spectral data and the Chl-a concentration of the water body,and the variable-order spectral dataset was constructed.The Partial Least Squares(PLS)model was used to screen the optimal features to construct the dataset,which was divided into modeling set and verification set according to the ratio of 7∶3,and the support vector machines(SVM)and deep forest(DF)models were used to establish the water Chl-a concentration inversion model.It is also compared with the model constructed using the original data and the model constructed by common dimensionality reduction methods.The results show that FOD technology can reduce hyperspectral noise,mine potential spectral information to a certain extent,and improve the correlation between hyperspectral reflectance and the concentration of Chl-a in water.Compared with the Chl-a inversion model established by using the original data and PCA dimension reduction,the R ^(2) of the water Chl-a concentration inversion model established by FOD combined with PLS first screening features was increased,and the mean square error(MSE)and mean absolute error(MAE)were reduced.DF has a higher degree of fitting and prediction accuracy than the other three models,with R ^(2)=0.96,MSE=0.51μg·L^(-1),and MAE=0.64μg·L^(-1).The validation set R ^(2)=0.89,MSE=0.69μg·L^(-1),MAE=0.64μg·L^(-1).In general,it is feasible to establish a water Chl-a concentration inversion model based on the variable-order spectral dataset after FOD recombination and the preferred features of PLS.Comparative analysis of the

关 键 词：叶绿素a反演 分数阶微分 偏最小二乘 深度森林 高光谱 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]