Generative Adversarial Network for SAR-to-Optical Image Translation with Feature Cross-Fusion Inference

Wei, Juan; Zou, Huanxin; Li, Sun; Cao, Xu; Li, Meilin; He, Shitian; Liu, Shuo

doi:10.1109/igarss46834.2022.9884166

Cited by 3 publications

(1 citation statement)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nevertheless, they are more difficult to interpret by human vision than optical images. Therefore, many works have been proposed to achieve SARto-optical (S2O) image translation [16][17][18][19][20][21][22][23] and to improve the readability of SAR images. However, geometric distortions in SAR images [24,25] are unavoidable, resulting from the special imaging mechanism of SAR images.…”

Section: Introductionmentioning

confidence: 99%

SAR-to-Optical Image Translation via an Interpretable Network

Zhang,

Zhang

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

Synthetic aperture radar (SAR) is prevalent in the remote sensing field but is difficult to interpret by human visual perception. Recently, SAR-to-optical (S2O) image conversion methods have provided a prospective solution. However, since there is a substantial domain difference between optical and SAR images, they suffer from low image quality and geometric distortion in the produced optical images. Motivated by the analogy between pixels during the S2O image translation and molecules in a heat field, a thermodynamics-inspired network for SAR-to-optical image translation (S2O-TDN) is proposed in this paper. Specifically, we design a third-order finite difference (TFD) residual structure in light of the TFD equation of thermodynamics, which allows us to efficiently extract inter-domain invariant features and facilitate the learning of nonlinear translation mapping. In addition, we exploit the first law of thermodynamics (FLT) to devise an FLT-guided branch that promotes the state transition of the feature values from an unstable diffusion state to a stable one, aiming to regularize the feature diffusion and preserve image structures during S2O image translation. S2O-TDN follows an explicit design principle derived from thermodynamic theory and enjoys the advantage of explainability. Experiments on the public SEN1-2 dataset show the advantages of the proposed S2O-TDN over the current methods with more delicate textures and higher quantitative results.

show abstract