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

Highly noise-resistant single face automatic super-resolution method

A super-resolution, high-resistance technology, applied in the fields of instruments, character and pattern recognition, computer parts, etc., can solve the problems of complex calculation method and processing process, reduce the effect of image restoration, slow calculation speed, etc., and achieve the recovery effect. Clear and natural, avoids sub-block search steps, and has a high degree of automation

Active Publication Date: 2012-09-26
ZHEJIANG ICARE VISION TECH
View PDF3 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Existing technologies such as the patent No. 200910056694X "Facial Image Super-resolution Processing Method Fused with Global Features and Local Information", this patent mainly constructs an image pyramid to represent the corresponding relationship between low frequency and high frequency, restores the global image, and then Forecasting residual items, the calculation method and processing process of this method are complicated, the calculation speed is slow, and the reconstruction efficiency is low
Another patent No. 200810232778X is "a method for super-resolution of face images reconstructed by using positional image blocks". How to achieve the purpose of anti-noise and anti-occlusion of the image, which reduces the effect of image restoration

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
  • Highly noise-resistant single face automatic super-resolution method
  • Highly noise-resistant single face automatic super-resolution method
  • Highly noise-resistant single face automatic super-resolution method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0016] The technical solutions of the present invention will be further specifically described below through examples.

[0017] Such as Figure 1-Figure 3 As shown, the present invention is specifically divided into two main processes of training and reconstruction. The training includes basic face model training and appendage model training. The basic face model training is to use the ASM algorithm to automatically calibrate the face of the high-resolution basic face database. Feature points, and through geometric transformation including affine transformation, projective transformation, polynomial transformation and triangulation transformation, reconstruct the face to a fixed and aligned basic face library, and then use the transformation to cut out the selected face library, and use PCA The algorithm is trained to obtain the face decomposition base sequence; after obtaining the basic face database, the face is divided into eyes, nose and mouth according to key areas, and t...

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 a highly noise-resistant single face automatic super-resolution method. The method of the invention is characterized by comprising two main processes of training and reconstruction. The training process comprises a basic face model training and an adjunct model training. The reconstruction process comprises the following steps: a face image is input; face feature points are calibrated by an ASM algorithm and subjected to geometric transformation so that a registered face image is obtained; a reconstructed basic face is obtained by means of an iterative method according to an algorithm of formula (1) min(W*||A*B*X-I[L]||<2>); current adjuncts are obtained by subtracting the reconstructed basic face from the input face image; reconstructed adjuncts are obtained by reconstruction by applying the algorithm of formula (1) to the adjuncts; and a final reconstructed face image is obtained by adding the reconstructed adjuncts and the registered face image. The super-resolution method of the invention has abilities of strong-noise resistance and occlusion resistance, provides a process with a high automation degree and a fast calculating speed, and enables a clear and natural recovery effect complying with common face structures, wherein the face component is replaceable so as to output multiple effects.

Description

technical field [0001] The invention belongs to the field of digital image processing, in particular to an automatic super-resolution method for a single human face with high noise resistance. Background technique [0002] Super-resolution refers to reconstructing one or more frames of high-resolution images from one or more frames of low-resolution images. Existing technologies such as the patent No. 200910056694X "Facial Image Super-resolution Processing Method Fused with Global Features and Local Information", this patent mainly constructs an image pyramid to represent the corresponding relationship between low frequency and high frequency, restores the global image, and then For the prediction residual item, the calculation method and processing process of this method are complicated, the calculation speed is slow, and the reconstruction efficiency is low. Another patent No. 200810232778X is "a method for super-resolution of face images reconstructed by using positional...

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 Applications(China)
IPC IPC(8): G06K9/00G06K9/62
Inventor 尚凌辉王百超高勇刘嘉
Owner ZHEJIANG ICARE VISION TECH
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