Image super-resolution reconstruction method combining depth supervision self-coding and perception iteration back projection

Abstract

The invention provides an image super-resolution reconstruction method combining depth supervision self-coding and perception iteration back projection. Compared with an existing method, a reconstruction model is directly trained, a low-resolution image is input into the trained reconstruction model to directly obtain a super-resolution image, and the reconstruction model cannot be adjusted afterbeing trained. According to the method, the degradation process from the super-resolution image to the low-resolution image is regarded as encoding, and the reconstruction process from the low-resolution image to the super-resolution image is regarded as decoding, so that an encoder reflecting an image complex degradation model is trained. According to the method, a bicubic interpolation image isused as an iteration initial value of the super-resolution image, a degenerated image of the super-resolution image generated by each iteration is obtained by using a trained encoder, the degeneratedimage is compared with an actual low-resolution image to obtain perception loss, and then the super-resolution image is updated by using the perception loss. According to the method, interference of large margin such as blur, jitter, noise and the like can be eliminated, and a high-resolution image can be reconstructed.

Description

technical field [0001] The invention belongs to the field of image processing and is mainly used for single image super-resolution reconstruction. [0002] technical background [0003] Image super-resolution reconstruction (Super-Resoluion, SR) is a research hotspot in the field of computer vision at present. Image, eliminate the irreversible degradation in the process of acquisition, transmission and storage, and reconstruct a clear and complete high-resolution image. Super-resolution reconstruction has a wide range of application scenarios in smart cities, big data medical care, multimedia social networking, autonomous driving and other fields, and is a very important digital image processing technology. Current image super-resolution reconstruction techniques include image interpolation methods, neighborhood embedding methods, sparse coding methods, and deep learning methods. These methods all presuppose the degradation relationship between bicubic interpolation and dow...

AI Technical Summary

Problems solved by technology

[0004] The invention solves the problem of image super-resolution reconstruction under complex degradation models such as noise, blur, compression, and downsampling, and proposes a new image super-resolution reconstruction method

Images