Image super-resolution implementation algorithm based on information filtering network

An information filtering and super-resolution technology, applied in the field of image processing, can solve the problems of blurred image outlines or lines, difficulty, poor image fusion effect, etc., to improve clarity, improve visual effects, and eliminate irrelevant information.

Pending Publication Date: 2020-06-16
海南鸿达盛创网络信息科技有限公司
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] Image super-resolution technology refers to the technology of recovering high-resolution images from one or more low-resolution images or image sequences. Its core idea is to exchange time resolution for higher spatial resolution. At present, there are mainly reconstruction-based There are two types of super-resolution methods and learning-based super-resolution methods, but they need to establish a degraded model that conforms to the actual imaging system and imaging conditions and perform accurate sub-pixel motion estimation on image sequences, which are difficult to achieve in actual processing. is very difficult and does not make full use of the prior information of the image
[0003] A Chinese patent discloses an image super-resolution algorithm based on information filtering network (authorized announcement number CN110223224A). This patented technology has a relatively small number of filters per layer and uses group convolution, so the network algorithm has a fast execution speed. It has the advantage of being fast, but its image fusion effect is not good, it cannot effectively eliminate irrelevant information in the image, and it is easy to make the outline or lines of the image blurred, which cannot improve the visual effect of the image very well.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image super-resolution implementation algorithm based on information filtering network
  • Image super-resolution implementation algorithm based on information filtering network
  • Image super-resolution implementation algorithm based on information filtering network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] see figure 1 , in an embodiment of the present invention, an image super-resolution implementation algorithm based on an information filtering network, including a contrast metric model, a brightness metric model, a normalization technique, a Gaussian pyramid, and a Laplacian pyramid;

[0031] Contrast metric model and brightness metric model are used to construct scalar weight maps; normalization techniques are used to calculate the mean and variance of the feature in the mini-batch, then subtract the mean and divide the feature by its mini-batch standard deviation; Gaussian pyramid is used for The image is zoomed in multiples, and the Laplacian pyramid is used to calculate the difference of the Gaussian pyramid, to obtain the high-frequency information of the image, and to restore and reconstruct the image.

[0032] Preferably, its implementation method comprises the following steps:

[0033] S1. Define the contrast metric model and brightness metric model of each in...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to the technical field of image processing, and discloses an image super-resolution implementation algorithm based on an information filtering network. The algorithm comprises acontrast measurement model, a brightness measurement model, a normalization technology, a Gaussian pyramid and a Laplace pyramid. The contrast measurement model and the brightness measurement model are used for constructing scalar weight mapping, and the Laplace pyramid is used for calculating the difference value of the Gaussian pyramid, obtaining the high-frequency information of the image and carrying out the restoration and reconstruction of the image. According to the invention, scalar weight mapping is constructed through the contrast measurement model and the brightness measurement model of the image, therefore, the bright color and the region of interest can obtain a large weight, the low contrast obtains smaller weight, better image fusion is facilitated, and the fusion pyramid isconstructed through the Gaussian pyramid and the Laplace pyramid, so that irrelevant information in the image can be effectively eliminated, the definition of image contour or line change is improved, and the visual effect of the image can be well improved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image super-resolution realization algorithm based on an information filtering network. Background technique [0002] Image super-resolution technology refers to the technology of recovering high-resolution images from one or more low-resolution images or image sequences. Its core idea is to exchange time resolution for higher spatial resolution. At present, there are mainly reconstruction-based There are two types of super-resolution methods and learning-based super-resolution methods, but they need to establish a degraded model that conforms to the actual imaging system and imaging conditions and perform accurate sub-pixel motion estimation on image sequences, which are difficult to achieve in actual processing. It is very difficult and does not make full use of the prior information of the image. [0003] A Chinese patent discloses an image super-resolution algori...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & AuthorityApplications(China)
IPC IPC(8): G06T3/40G06T5/00G06T5/50G06T7/90
CPCG06T3/4076G06T5/50G06T7/90G06T2207/20016G06T2207/20221G06T5/00G06T5/70
Inventor阳洪求
Owner海南鸿达盛创网络信息科技有限公司