Image restoration method based on generative adversarial neural network

Abstract

The invention discloses an image restoration method based on a generative adversarial neural network, and the method employs four parts: a generative network, a global discrimination network, a localdiscrimination network, and an LSTM neural network. And the two discriminators are mainly used for ensuring that the repaired missing area can keep consistent with the surrounding area. The LSTM neural network is mainly used for repairing damaged images by stages. The algorithm comprises a data preprocessing module, a model training module and an image restoration module, and is mainly used for performing a semantic restoration task on a large-area missing image so as to reconstruct a complete and vivid image conforming to human eye senses.

Description

technical field [0001] The invention relates to an image restoration method based on a generative confrontational neural network, which belongs to the field of image semantic restoration. Background technique [0002] As an important information carrier of the objective world, images are the main source and means for human beings to obtain and identify external information. High-quality images can bring people richer information and content, and even give people a beautiful viewing experience. However, in real life, it is not guaranteed that all acquired images are of high quality. In the process of image collection, transmission, storage, etc., after some image processing operations, image information will often be lost. decline in quality. For example, in the process of image transmission, due to channel bandwidth limitation or channel damage, information may be lost in the transmitted image. In order to improve the image quality and ensure the effectiveness of informat...

AI Technical Summary

Problems solved by technology

However, on the one hand, when inpainting an image, it is usually difficult to find a general inpainting law due to the complex texture structure and uncertain noise of the missing area and background, so it is very difficult to directly build a mathematical model for inpainting

On the other hand, since image inpainting is an ill-conditioned inverse problem, it reconstructs the information of the missing area through the learning, understanding and inference of the image perception process according to the acquired part of the image information. However, the reconstruction result is not unique. Finding the optimal inpainting result is also a challenging problem

Images