A single image super-resolution method based on reversible network

Abstract

The invention discloses a single image super-resolution method based on a reversible network, which belongs to the field of image processing. The method introduces a reversible network to construct the network structure of the super-resolution model, and utilizes the reversible nature of the reversible network to realize the mutual mapping of the high-resolution image space and the low-resolution image space, from the low-resolution and high-resolution two The direction optimizes the super-resolution process, which solves the problem that other deep learning-based super-resolution methods cannot effectively use the interdependence between high-resolution and low-resolution images, thereby improving the model's ability to perform image super-resolution . Also by introducing singular value decomposition to initialize the weight matrix of the 1×1 reversible convolutional layer, the propagation speed of the inverse process of the 1×1 reversible convolutional layer is improved; the method of this application can effectively realize the super-resolution process of a single image, using Low-resolution images generate super-resolution images with good texture details and visual effects.

Description

Technical field[0001]The invention relates to a single image super-resolution method based on a reversible network, and belongs to the field of image processing.Background technique[0002]Image Super Resolution refers to the restoration of a high-resolution image from a low-resolution image or image sequence. Single image super resolution (Single image super resolution, SISR) is to establish a mapping between low-resolution images and high-resolution images, and generate super-resolution images from the input low-resolution images. The current super-resolution methods mainly include methods based on interpolation, reconstruction, and learning. Common interpolation methods include bilinear interpolation, bicubic interpolation, etc., but the reconstructed image obtained by interpolation is prone to blur, sawtooth, and lack of texture details. Reconstruction-based methods are based on the use of low-resolution images as constraints, combined with prior knowledge of the image to reconstr...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem that the current existing technology does not utilize the interdependence between the low-resolution image and the high-resolution image, and cannot use the mutual information between the two images more effectively, thus affecting the effect of model image super-resolution, The present invention provides a single image super-resolution method based on a reversible network. The method uses a reversible network to construct a network model for super-resolution, and then inputs a low-resolution image to one end of the network model to generate a super-resolution image. Input the high-resolution image to the other end of the network model to generate a low-resolution reconstruction image, and use the generated super-resolution image and the difference between the low-resolution reconstruction image and the real high-resolution image and low-resolution image to design The optimized objective function updates the parameters of the network model by minimizing the value of the objective function to improve the super-resolution capability of the network model

Images