Sentinel-1 Dual-Polarization SAR Images Despeckling Network Based on Unsupervised Learning
Jie Li,Liupeng Lin,Mange He,Jiang He,Qiangqiang Yuan,Huanfeng Shen
DOI: https://doi.org/10.1109/tgrs.2024.3404405
IF: 8.2
2024-06-12
IEEE Transactions on Geoscience and Remote Sensing
Abstract:Supervised deep learning despeckling methods usually use optical images to simulate multiplicative noise for training. However, due to the different imaging mechanisms of optical images and synthetic aperture radar (SAR) images, the data characteristics of the two are significantly different, resulting in poor generalization performance of the model trained through the above form. Besides, the existing deep learning models do not fully consider the physical scattering mechanism, which causes the loss of polarization information. To solve those problems, an unsupervised deep learning method is proposed for dual-polarization SAR image despeckling. Under this framework, we combine the dual-polarization SAR covariance matrix and polarization decomposition information to construct a dual-branch SAR image despeckling network (DSDN). The residual channel and the spatial attention mechanism are embedded to calibrate the polarization and spatial feature maps. The cross-attention (CroA) mechanism is designed to mine the association of feature maps before and after denoising. Besides, the dual-branch joint loss function is proposed to constrain the training process. Spatial information experiments and polarization information experiments indicate that, compared with the existing state-of-the-art SAR despeckling methods, the proposed method can effectively remove the coherent speckle noise of dual-polarization SAR images and can better preserve the polarization information. Codes are available at https://github.com/LiupengLin/DSDN.
imaging science & photographic technology,remote sensing,engineering, electrical & electronic,geochemistry & geophysics