Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Image super-resolution method based on improved non-local constraints and local self-similarity

A self-similar, non-local technology, applied in the field of image processing, it can solve the problems of blurred edges of high-resolution images, prone to errors, poor visual effects, etc., and achieve the effect of more image details and sharpening image edges.

Active Publication Date: 2016-04-13
XIDIAN UNIV
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, because this method only searches for a similar high-frequency image block, it depends too much on the matching criterion, which is prone to errors, and the obtained high-resolution image has blurred edges and poor visual effects.

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 method based on improved non-local constraints and local self-similarity
  • Image super-resolution method based on improved non-local constraints and local self-similarity
  • Image super-resolution method based on improved non-local constraints and local self-similarity

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] refer to figure 1 , the implementation steps of the present invention are as follows:

[0031] Step 1, input a low-resolution image X, and interpolate and enlarge it.

[0032] Enter as image 3 For the low-resolution image X shown, set the magnification factor λ=1.25, perform bicubic interpolation and amplification on the low-resolution image X, and obtain the pre-amplified image Y o

[0033] Step 2, filter the low-resolution image X.

[0034] Use a Gaussian high-pass filter to filter the input low-resolution image X and decompose it into high-frequency components X h and the low frequency component X o .

[0035] Step 3, using the high frequency component X h , low frequency component X o and the prescaled image Y o , perform super-resolution reconstruction on the input low-resolution image X.

[0036] 3a) In the low frequency component X o Extract low frequency image blocks from X o j ,j=1,...,M, M is the number of low-frequency image blocks, in the pre-a...

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 discloses an image super-resolution method based on improved non-local restriction and local self similarity. The method includes the steps that 1, bicubic interpolation amplification and high-pass filtering are conducted on a low-resolution image X to acquire a pre-amplified image, low-frequency components and high-frequency components; 2, image blocks are extracted from the pre-amplified image and the low-frequency components; 3, K clustering is conducted on the low-frequency image blocks, the pre-amplified image blocks are compared with the center of clustering of each cluster to find out the most similar cluster, and then three similar low-frequency image blocks are found out in the most similar cluster; 4, corresponding high-frequency image blocks are found out according to the similar low-frequency image blocks, and non-local weighing is conducted on the high-frequency image blocks to acquire a reconstructed preliminary high-resolution image; 5, the reconstructed preliminary high-resolution image is used ad a next input image, and the first step, the second step, the third step and the fourth step are repeatedly executed to acquire a final high-resolution image. By the adoption of the method, in the super-resolution process of the images, the edges of the images can be sharpened, and high-frequency details of the images can be well restored.

Description

technical field [0001] The invention belongs to the technical field of image processing, and specifically relates to an image resolution method, which can be used to super-resolve an input low-resolution image to obtain a high-resolution image. Background technique [0002] Image super-resolution technology is a discipline that improves image clarity and suppresses noise through various technical means in order to obtain more accurate image information. It is an important and challenging research content in image processing. For the image super-resolution problem, researchers have proposed many methods. [0003] In 2008, Yang et al. proposed an image super-resolution reconstruction method based on example learning, see J. Yang, J. Wright, T. Huang, and Y. Ma, "Image super-resolution ass parse representation of raw image patches", inProc.IEEEConf.Comput.Vis. PatternRecognit., 2008, pp.1-8. The basic idea of ​​this method is to randomly select some blocks from some high-reso...

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 & Authority Patents(China)
IPC IPC(8): G06T5/00
CPCG06T3/4053
Inventor 王爽焦李成张阳马文萍马晶晶刘红英
Owner XIDIAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products