Single-image super-resolution reconstruction algorithm based on multi-scale residual error learning network

Abstract

The invention relates to a single-image super-resolution reconstruction algorithm based on a multi-scale residual error learning network. In recent years, the convolutional neural network is widely applied to many visual tasks, and particularly, remarkable results are obtained in the field of single-image super-resolution reconstruction. Similarly, multi-scale feature extraction also achieves consistent performance improvement in the field. However, in the prior art, multi-scale features are extracted in a layered mode mostly, and with the increase of the depth and width of a network, the calculation complexity and the consumption of a memory can be greatly improved. In order to solve the problem, the invention provides a compact multi-scale residual error learning network, i.e., representing multi-scale characteristics in a residual error block. The model is composed of a feature extraction block, a multi-scale information block and a reconstruction block. In addition, due to the factthat the number of network layers is small and group convolution is used, the network has the advantage of being high in execution speed. Experimental results show that the method is superior to an existing method in time and performance.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a single image super-resolution reconstruction algorithm based on a multi-scale residual learning network. Background technique [0002] Image super-resolution reconstruction, especially single image reconstruction, has attracted more and more attention. The main task of super-resolution reconstruction is to reconstruct high-resolution images from low-resolution images with reasonable priors. It is worth noting that this is an ill-posed problem, because there may be different mapping relationships between high-resolution and low-resolution images. Therefore, these specific mapping relationships need to be obtained by learning from large image datasets. Traditional methods based on external examples can effectively solve this problem, that is, learning compact dictionaries or manifold spaces from external datasets to represent this mapping relationship, such as nearest ...

AI Technical Summary

Problems solved by technology

In order to obtain better reconstruction results, the depth and width of the network are continuously increased, which greatly increases the computational complexity and memory consumption

Images