基于高光谱数据波段选择的茅洲河悬浮物浓度反演  

Inversion of Suspended Particulate Matter Concentration in Maozhou River Based on Band Selection of Hyperspectral Data

作　　者：陈仲锴 厉小润[1] 赵辽英[2] Chen Zhongkai;Li Xiaorun;Zhao Liaoying(College of Electrical Engineering,Zhejiang University,Hangzhou,Zhejiang 310027,China;Institute of Computer Application Technology t Hangzhou Dianzi University,Hangzhou,Zhejiang 310018,China)

机构地区：[1]浙江大学电气工程学院,浙江杭州310027 [2]杭州电子科技大学计算机应用技术研究所,浙江杭州310018

出　　处：《激光与光电子学进展》2021年第20期27-35,共9页Laser & Optoelectronics Progress

摘　　要：针对基于高光谱数据的悬浮物浓度反演问题,提出一种利用预训练神经网络(PNN)进行有监督波段选择的方法,并使用随机森林和神经网络建立悬浮物浓度反演模型。PNN算法需进行多次重复实验以获得低噪声且充分的波段重要性表达,在每次实验时选取适当数量的波段作为神经网络输入数据的特征,训练一个神经网络,并利用最后一期训练时第一层权重的L1范数、L2范数和ReLU(Rectified Linear Unit)函数来表示波段的重要性。实验结果表明,相比其他常用的波段选择算法,使用L1范数、L2范数的PNN算法能够得到信息量更大的波段集合,且其用于悬浮物浓度反演时,能够获得更优的效果。Aiming at the inversion of suspended particulate matter(SPM)concentration using hyperspectral data,this paper proposes a supervised band selection method based on pre-trained neural networks(PNNs),and employs the random forest and neural network to establish an inversion model of SPM concentration.The PNN method needs to perform multiple repeated experiments to obtain sufficient and low-noise expression of band importance.In each experiment,an appropriate number of bands is selected as the features of input data of neural networks.Then,we train a neural network and obtain weights of the first layer in the last training epoch.Finally,we use the L1 norm,L2 norm,and ReLU(Rectified Linear Unit)function of the weights to represent the importance of the bands.The experiment results show that the PNN method using L1 norm and L2 norm can obtain a more informative band set,and perform better when used for SPM concentration inversion.

关 键 词：大气光学 高光谱数据 波段选择 预训练神经网络 悬浮物浓度反演 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]