Toward Blind-Adaptive Remote Sensing Image Restoration

• Published in: IEEE transactions on geoscience and remote sensing Vol. 61; pp. 1 - 12
• Main Authors: Liu, Maomei, Tang, Lei, Fan, Lijia, Zhong, Sheng, Luo, Hangzai, Peng, Jinye
• Format: Journal Article
• Language: English
• Published: New York IEEE 2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptive control; Artificial neural networks; Context; Control systems design; Convolution; Convolutional neural network (CNN); Image coding; Image compression; Image restoration; Information processing; JPEG-LS compression; Model accuracy; Modules; Neural networks; Q-factor; Receptive field; Remote sensing; remote sensing image restoration; Restoration; Sensors; Task analysis; Transform coding
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2023.3318250

While deep convolutional neural networks (CNNs) have substantially boosted the performance of low-level vision tasks, they remain largely underexplored in CNN-based remote sensing image restoration. This article studies the JPEG-LS-compressed remote sensing image restoration that faces the following problems. It requires a tradeoff in preserving local context information and expanding spatial receptive fields. It needs blind restoration while achieving flexible performance. To this end, we propose a blind-adaptive restoration network, called TBANet, which integrates three modules into an end-to-end network to remedy these problems separately. Specifically, we build a scale-invariant wise-skip (SIWS) ResNet as the baseline to extract more context information. We present a receptive field expansion module using scalewise convolution for removing banding artifacts. We design a blind-adaptive controller to provide a deterministic result meanwhile meeting the needs of the user's preference. In experiments, we compare the restoration accuracy among our model and many different variants of restoration methods on our collected remote sensing image dataset. The proposed network achieves superior performance against state-of-the-art methods in terms of both quantitative metrics and visual quality. Code and models are available at: https://github.com/lmmhh/TBANet .