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

Image restoration method based on multi-scale structure block

A repair method and multi-scale technology, applied in the field of image processing, can solve problems such as distortion, and achieve the effect of suppressing artificial effects

Active Publication Date: 2016-10-12
XIDIAN UNIV
View PDF2 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to propose an image restoration method based on multi-scale structural blocks to analyze the structural information of image blocks at different scales to calculate Priority, which improves the accuracy of the priority judgment of the texture area (including irregular texture), and at the same time, by prioritizing the repair of the structure area whose boundary is in the middle, the structure information can be maintained and the distortion of the structure can be reduced

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 restoration method based on multi-scale structure block
  • Image restoration method based on multi-scale structure block
  • Image restoration method based on multi-scale structure block

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] The present invention is an image restoration method based on multi-scale structural blocks, see figure 1 , including the following steps:

[0034] (1) Input the image I to be repaired, see Figure 4 , determine the area to be repaired Ω and the edge of the area to be repaired

[0035] (2) For any pixel p on the edge of the area to be repaired, Calculate the priority of all edge pixels in the area to be repaired according to the multi-scale structure information of the edge pixels p in the area to be repaired;

[0036] The multi-scale structure information of the edge pixel p in the area to be repaired is composed of two parts; the first part of the multi-scale structure information is the inconsistency measurement of the edge pixel p in the area to be repaired on different scales, which measures the location of the edge pixel p in the area to be repaired. Structural information within the region. The specific method is to calculate the inconsistency measures of ...

Embodiment 2

[0044] An image restoration method based on multi-scale structure blocks is the same as embodiment 1, and according to the multi-scale structure information in step (2), embodiment 2 is aimed at Figure 7 Calculate the priority of all the edge pixels in the region to be repaired. Calculating priority through multi-scale structural information includes the following steps:

[0045] (2.a) For an image block whose center is at the edge pixels of the area to be repaired, on the block scale f*f, divide the block along the direction, and the specific direction is to divide the block evenly along the direction j, where the direction j is defined by the plane coordinate system Orthogonal orientation groups in . This example targets Figure 7 , set the direction j as: in the plane polar coordinate system, the four directions of θ=0°, 90°, 45°, -45°, these four directions correspond to the directions j=1, 2, 3, 4 respectively; see figure 2 .

[0046] For any pixel p on the edge of ...

Embodiment 3

[0072] A kind of image restoration method based on multi-scale structure block is the same as embodiment 1-2, in step (3), embodiment 3 is aimed at Figure 10 For all the edge pixels of the area to be repaired, set the edge pixel of the area to be repaired with the highest priority as q; then according to the priority of pixel q, adaptively determine the block size f*f of the block to be repaired for pixel q.

[0073] The specific methods include: according to the inconsistency measurement Dif(Ψ q ) (its block scale is f*f) and scale consistency Sca(q), the pixel q is divided into: structural edge pixels, random edge pixels or general edge pixels.

[0074](3.a) If Dif(Ψ q )>Tresh2 and Sca(q)>Tresh1; then the pixel q belongs to the edge pixel of the structure; let the block to be repaired be the edge block ψ with the highest priority q , f in its block size: f=f1, where f1>f;

[0075] (3.b) If Dif(Ψ q )q , f in its block size: f=f2, where f2<f;

[0076] (3.c) Others, the p...

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 restoration method based on a multi-scale structure block, and mainly solves the technical problem that distortion is easy to occur when a larger structure region is restored in the prior art. The method comprises the following realization processes: for edge pixels of regions to be restored, calculating priority through inconsistency measurement and scale invariance measurement under different scales; for the edge pixel of the region to be restored, the current priority of which is the highest, determining the size of a block to be restored adaptively according to the priority thereof; restoring the block to be restored through nonlocal reconstruction technology; and traversing the regions to be restored in the image, and repeating the steps above until finishing restoration. Structure information of image blocks in different scales is utilized, and effective structure blocks are selected for restoration, thereby improving irregular texture and random content priority judgment, and keeping image texture and structure consistency and suppressing artificial effect. A better visual effect and data index image restoration result is obtained, and the method can be used for broken image restoration and target object removal.

Description

technical field [0001] The invention belongs to the technical field of image processing, and mainly relates to image restoration, in particular to an image restoration method based on multi-scale structural blocks, which can be used for restoration of natural images. Background technique [0002] Digital image restoration is an important part of image restoration research. Its purpose is to automatically recover lost information based on the existing information of the image. Repair of scratches and missing information etc. Therefore, digital image restoration occupies a pivotal position in the field of digital image processing and has become one of the most basic technologies in this field. However, since the image of the area to be repaired is unknown, there are many technical difficulties to be solved in digital image restoration. [0003] Existing inpainting methods can be roughly divided into two categories, one is the inpainting method based on diffusion, and the oth...

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
IPC IPC(8): G06T5/00
CPCG06T2207/10004G06T5/77
Inventor 钟桦焦李成胡雪纯田小林缑水平
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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com