Convolution Neural Network based lossy compression of hyperspectral images

作者：

Highlights：

• Use of CNN for compression of HSI.

• Autoencoder model extracts the features in a lower dimension tensor.

• Proposed algorithm outperforms existing techniques.

• Application-oriented analysis of the reconstructed image.

摘要

•Use of CNN for compression of HSI.•Autoencoder model extracts the features in a lower dimension tensor.•Proposed algorithm outperforms existing techniques.•Application-oriented analysis of the reconstructed image.

论文关键词：Autoencoder,Lossy compression,Convolution Neural Network,Hyperspectral images,Learning based compression

论文评审过程：Received 23 April 2020, Revised 6 February 2021, Accepted 28 March 2021, Available online 7 April 2021, Version of Record 15 April 2021.

论文官网地址：https://doi.org/10.1016/j.image.2021.116255