AIF_Pix2Pix: An Ocean Front Airborne Infrared Image Generation Model

We present an airborne infrared front Pix2Pix (AIF_Pix2Pix) model, which generates airborne infrared images of ocean fronts via inputs processed by a front region mask and region brightness perturbations. The improvements applied to this model include reducing the weight of the network, modifying its loss function, and adding data augmentation on the basis of the Pix2Pix framework. With a parameter reduction of 88.6%, the AIF_Pix2Pix model demonstrates superior performance to that of the existing approaches: the structural similarity index measure (SSIM) and peak signal-to-noise ratio (PSNR) increase by 2.6% and 6.2%, respectively, and the root mean squared error (RMSE) decreases by 31.4%. In particular, the generated ocean fronts are highly consistent with real data in terms of their positions and intensities, with a reduction of 32.8% in the standard deviation of the gradient differences. Additionally, we provide the effectiveness results generated by AIF_Pix2Pix from actual airborne infrared imagery.