Restoration of turbulence-degraded images using the modified convolutional neural network

Atmospheric turbulence can severely degrade images obtained by remote imaging systems. The most noticeable effects of atmospheric turbulence on imaging are space-time distortions and various blurs. In this paper, to restore a single high-quality image from degraded images distorted by atmospheric turbulence, we use an end-to-end fusion subnetwork with smoothed dilated convolution and multiscale mapping modules. The proposed model can be embedded in high-level vision tasks and considered a "real-time" task, as it takes less than a second to resume an image. Extensive experiments on both simulated and real-world data prove that our approach outperforms previous state-of-the-art approaches in both quantitative and qualitative terms.