Two-stage deep neural network image restoration method based on super-resolution technology

The invention discloses a two-stage deep neural network image restoration method based on a super-resolution technology, and relates to the technical field of image restoration. The two-stage deep neural network image restoration method based on the super-resolution technology comprises the following steps: acquiring a damaged image; the damaged image is repaired based on the constructed image repairing network, the image repairing network comprises a rough repairing network and a refined repairing network, the rough repairing network is used for preliminarily repairing the damaged image to obtain a preliminarily repaired image, and the refined repairing network is used for finely repairing the preliminarily repaired image to obtain a refined repaired image; according to the invention, the problem of how to recover high-resolution information from a low-resolution image is solved. 本发明公开了一种基于超分辨率技术的双阶段深度神经网络图像修复方法，涉及图像修复技术领域。该基于超分辨率技术的双阶段深度神经网络图像修复方法，包括以下步骤：获取受损图像；基于构建的图像修复网络对受损图像进行修复，所述图像修复网络包括粗略修复网络和精细化修复网络，