Physics-Driven Turbulence Image Restoration with Stochastic Refinement
This addresses a critical problem in optical imaging systems by improving generalization to real-world turbulence conditions, though it is incremental as it builds on existing deep-learning and simulation methods.
The paper tackles image distortion from atmospheric turbulence in long-range optical imaging by proposing a Physics-integrated Restoration Network (PiRN) that incorporates a physics-based simulator during training to separate stochastic degradation from the underlying image, and introduces PiRN with Stochastic Refinement (PiRN-SR) to enhance perceptual quality, achieving state-of-the-art restoration in pixel-wise accuracy and perceptual quality.
Image distortion by atmospheric turbulence is a stochastic degradation, which is a critical problem in long-range optical imaging systems. A number of research has been conducted during the past decades, including model-based and emerging deep-learning solutions with the help of synthetic data. Although fast and physics-grounded simulation tools have been introduced to help the deep-learning models adapt to real-world turbulence conditions recently, the training of such models only relies on the synthetic data and ground truth pairs. This paper proposes the Physics-integrated Restoration Network (PiRN) to bring the physics-based simulator directly into the training process to help the network to disentangle the stochasticity from the degradation and the underlying image. Furthermore, to overcome the ``average effect" introduced by deterministic models and the domain gap between the synthetic and real-world degradation, we further introduce PiRN with Stochastic Refinement (PiRN-SR) to boost its perceptual quality. Overall, our PiRN and PiRN-SR improve the generalization to real-world unknown turbulence conditions and provide a state-of-the-art restoration in both pixel-wise accuracy and perceptual quality. Our codes are available at \url{https://github.com/VITA-Group/PiRN}.