A Super-resolution Reconstruction Method of Blurred Image Based on Fractional Differential

Abstract

The invention provides a fuzzy image super-resolution reconstruction method based on fractional differential. The method comprises the steps of (1) carrying out Gauss fuzzy and down sampling processing on an original image set in order, and obtaining an input image sequence, (2) selecting one arbitrary frame of input image from the input image sequence to carry out bicubic interpolation amplification, and obtaining an original reference frame, (3) using an adaptive fractional differential algorithm to carrying out image enhancement processing on each frame of input image in the input image sequence and the original reference frame, (4) calculating a motion matrix between each frame of input image and the original reference frame through SIFT matching, and searching a point in each frame of input image corresponding to a point in the original reference frame, and (5) calculating a residual error between the point in the original reference frame and the corresponding point in each frame of input image, through residual error inverse iterative projection correction, and the pixel value of the point in the original reference frame is constantly adjusted until a preset condition is satisfied.

Description

technical field [0001] The invention relates to an image super-resolution reconstruction method, in particular to a fractional differential-based fuzzy image super-resolution reconstruction method. Background technique [0002] With the development of computer technology and machine vision technology, the method of image super-resolution has made great progress, and has important applications in many fields, such as multimedia, security monitoring and medical fields. [0003] At present, image super-resolution algorithms can be mainly divided into three categories: methods based on interpolation theory, methods based on learning, and methods based on reconstruction and enhancement theory. It will be accompanied by a large amount of information loss and blurring, which cannot normally achieve the goal in super-resolution of blurred images. Although the effect of image super-resolution based on learning is good, it relies heavily on external databases. Without a good and suff...

AI Technical Summary

Problems solved by technology

[0003] At present, image super-resolution algorithms can be mainly divided into three categories: methods based on interpolation theory, methods based on learning, and methods based on reconstruction and enhancement theory. It will be accompanied by a large amount of information loss and blurring, which cannot normally achieve the goal in super-resolution of blurred images

Although the effect of image super-resolution based on learning is good, it relies heavily on external databases. It cannot effectively achieve the goal without a good and sufficient database, and its calculations also limit its application.

However, most existing methods based on reconstruction or enhancement are not robust enough for super-resolution of blurry images.

Images