Classification of hyperspectral imagery using a fully complex-valued wavelet neural network with deep convolutional features

作者：

Highlights：

• A novel approach is proposed for the classification of hyperspectral imagery (HSI).

• A novel complex-valued attribute set is obtained for the HSI classification.

• Complexity analysis shows the applicability of the algorithm.

• Experimental results demonstrate the superiority and effectiveness of the method.

摘要

•A novel approach is proposed for the classification of hyperspectral imagery (HSI).•A novel complex-valued attribute set is obtained for the HSI classification.•Complexity analysis shows the applicability of the algorithm.•Experimental results demonstrate the superiority and effectiveness of the method.

论文关键词：Hyperspectral image classification,Complex-valued wavelet neural network (CVWNN),Convolutional neural network (CNN),Deep feature extraction

论文评审过程：Received 11 April 2020, Revised 14 September 2020, Accepted 9 February 2021, Available online 16 February 2021, Version of Record 26 February 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.114708