期刊
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
卷 33, 期 12, 页码 7251-7265出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TNNLS.2021.3084682
关键词
Superresolution; Tensors; Spatial resolution; Hyperspectral imaging; Computer architecture; Training; Learning systems; Attention module (AM); deep convolutional neural network (CNN); hyperspectral image (HSI) super-resolution; image fusion; pixelShuffle (PS)
类别
资金
- National Natural Science Foundation of China [61772003, 61702083, 12001446, 61876203]
- National Key Research and Development Program of China [2020YFA0714001]
- Key Projects of Applied Basic Research in Sichuan Province [2020YJ0216]
- Fundamental Research Funds for the Central Universities [JBK2102001]
- MIAI@Grenoble Alpes [ANR19-P3IA-0003]
The paper introduces a deep convolutional neural network architecture to fuse low-resolution HSI and high-resolution multispectral image for generating high-resolution HSI. By preserving spatial and spectral information using LR-HSI and HR-MSI, and utilizing attention and pixelShuffle modules for high-quality spatial details extraction, the proposed network achieves the best performance compared to recent HSI super-resolution approaches.
Hyperspectral images (HSIs) are of crucial importance in order to better understand features from a large number of spectral channels. Restricted by its inner imaging mechanism, the spatial resolution is often limited for HSIs. To alleviate this issue, in this work, we propose a simple and efficient architecture of deep convolutional neural networks to fuse a low-resolution HSI (LR-HSI) and a high-resolution multispectral image (HR-MSI), yielding a high-resolution HSI (HR-HSI). The network is designed to preserve both spatial and spectral information thanks to a new architecture based on: 1) the use of the LR-HSI at the HR-MSI's scale to get an output with satisfied spectral preservation and 2) the application of the attention and pixelShuffle modules to extract information, aiming to output high-quality spatial details. Finally, a plain mean squared error loss function is used to measure the performance during the training. Extensive experiments demonstrate that the proposed network architecture achieves the best performance (both qualitatively and quantitatively) compared with recent state-of-the-art HSI super-resolution approaches. Moreover, other significant advantages can be pointed out by the use of the proposed approach, such as a better network generalization ability, a limited computational burden, and the robustness with respect to the number of training samples. Please find the source code and pretrained models from https://liangjiandeng.github.io/Projects_Res/HSRnet_2021tnnls.html.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据