Dehaze-TGGAN: Transformer-Guide Generative Adversarial Networks With Spatial-Spectrum Attention for Unpaired Remote Sensing Dehazing
- 29 July 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 62 (01962892) , 1-20
Abstract
Satellite imagery plays a critical role in target detection. However, the quality and usability of optical remote sensing images can be severely compromised by atmospheric conditions, particularly haze, which significantly reduces the recognition accuracy of target detection algorithms such as ships. On the other hand, paired training data, i.e. the remote sensing data with or without fog at the same place, are difficult to obtain in real-world scenarios, leading to the failure of many existing dehazing methods. To deal with these issues, this paper proposes a Transformer-Guide CycleGAN framework Generative Adversarial Networks (Dehaze-TGGAN) incorporating an extra attention mechanism from frequency domain. Firstly, an SSA mechanism is proposed by using a 2-Dimensional Fast Fourier Transform (2D FFT) in the spatial domain, which enables the model to understand the relationships within the three-channel frequency domain information and to recover the spectral features of the hazy image through the spectrum encoder block. Then, a pre-training approach using Semi-Transparent Masks (STM), which can effectively simulate hazy conditions by adjusting the transparency of masks, is presented as a key strategy to accelerate the convergence rate. Finally, the applicability of the transformer architecture is extended by incorporating Total Variation Loss (TV Loss). The results of simulated and measured optical remote sensing data show that the recognition accuracy and the efficiency of the proposed algorithm are greatly improved.Keywords
Funding Information
- National Natural Science Foundation of China (62171379, 62271409)
This publication has 55 references indexed in Scilit:
- Ensemble learning-based method for maritime background subtraction in open sea environmentsComputer Vision and Image Understanding, 2023
- GLF-CR: SAR-enhanced cloud removal with global–local fusionISPRS Journal of Photogrammetry and Remote Sensing, 2022
- MSTFDN: Multi-scale transformer fusion dehazing networkApplied Intelligence, 2022
- Research on Remote Sensing Image De‐haze Based on GANJournal of Signal Processing Systems, 2021
- Cloud removal in Sentinel-2 imagery using a deep residual neural network and SAR-optical data fusionISPRS Journal of Photogrammetry and Remote Sensing, 2020
- Tracking Persons-of-Interest via Unsupervised Representation AdaptationInternational Journal of Computer Vision, 2019
- Attention-GAN for Object Transfiguration in Wild Images2018
- Single Image Dehazing via Multi-scale Convolutional Neural Networks2016
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015
- Haze detection and removal in high resolution satellite image with wavelet analysisIEEE Transactions on Geoscience and Remote Sensing, 2002