Super-resolution method of noise-containing image

Abstract

The invention relates to a super-resolution method of a noise-containing image. The method mainly comprises the following steps that: a low-resolution noise-containing image I is inputted; simple interpolation is performed on the image I so as to obtain the preliminary estimate L<0>k+1 of a low-resolution test image, and with the preliminary estimate L<0>k+1 of the low-resolution test image adopted as a basis, image multi-scale pyramids in a horizontal direction, a vertical direction and a diagonal direction are obtained, a low-resolution test image Lk+1 is obtained through using the image multi-scale pyramids; Gaussian blur operation is performed on the image I, and a blurred image Lk is obtained; and with image blocks adopted as units, the mapping function of the internal image blocks ofI and Lk is solved, the mapping function is applied to corresponding internal image blocks of Lk+1, so that high-resolution image blocks can be generated, and a high-resolution image is obtained. With the method of the invention adopted, noises can be well suppressed in a super-resolution process, and the details of the edge of the reconstructed image are more abundant, and the deformation of theimage is less.

Description

technical field [0001] The invention relates to the field of image super-resolution, in particular to a noise-containing image super-resolution method. Background technique [0002] Limited by the physical conditions of imaging devices (digital cameras, hand-held video cameras, etc.), the acquired digital images often have low resolution and contain noise. Among them, the resolution of a digital image is one of the important indicators to measure the quality of an image. The higher the resolution, the clearer the image and the richer the information it can provide. To a certain extent, the increase in image resolution is accompanied by an increase in the number of pixels. In order to improve the resolution of the image, most of the existing technical means estimate the newly added pixel information inside the high-resolution image through the pixels inside the obtained image (hereinafter referred to as "low-resolution image"). Such technical means are collectively referred...

AI Technical Summary

Problems solved by technology

[0007] In summary, how to solve the problem that the existing technology is easy to cause image noise amplification during the super-resolution process of digital images, so that it is difficult to deal with noisy images, has become an urgent problem in the field of image super-resolution

Images