Face image deblurring method based on deep learning

A face image and deep learning technology, applied in the field of image processing, can solve the problems of excessive image deblurring, inaccuracy, deblurring, etc., and achieve the effect of good effect, fast speed, and guaranteed robustness

Active Publication Date: 2020-11-24
HOHAI UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, these priors are estimated from limited observations and are not accurate enough
As a result, deblurred images are usually either under-deblurred (image remains blurry) or over-deblurred (image contains many artifacts)

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
  • Face image deblurring method based on deep learning
  • Face image deblurring method based on deep learning
  • Face image deblurring method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] The invention will be described in further detail below in conjunction with the accompanying drawings.

[0036] Such as figure 1 Shown is the flow chart of the present invention, which extracts the content information of the clear face image and the blurred face image respectively, and reconstructs the blurred and clear face images according to the obtained image content information. Introducing cycle consistency loss, the deblurred clear face image is re-blurred to reconstruct the original blurred image, and the blurred clear face image is transformed into the original clear image. Finally, the deblurred face image is generated by a clear image generator.

[0037] Such as figure 2 Shown is the deblurring framework of the present invention, S is a clear face image, b is a blurred face image, is the content encoder for blurred images and is a content encoder for sharp images, E b is the fuzzy encoder, G B is the image generator for the blurred image and is G S ...

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 a face image deblurring method based on deep learning, and belongs to the technical field of image processing. According to the invention, while the content structure of the original image is reserved, the content encoder and the blurred encoder are used for distinguishing the content of the blurred image from the blurred features, so that high-quality image deblurring is realized. According to the method, fuzzy branches and cyclic consistency loss are added into a framework, the robustness of the model is ensured by further restoring cyclic consistency, and the added perception loss is beneficial to removing unpractical artifacts from the fuzzy image and finally adding adversarial loss, so that the deblurring effect of the method is better and the speed is higher.

Description

technical field [0001] The invention relates to a face image deblurring method based on deep learning, belonging to the technical field of image processing. Background technique [0002] The subject of image deblurring has received widespread attention and has a long research history, not only because digital images have become an important way for people to obtain, exchange and understand information, but also because images contain a large amount of information that other signals cannot match. Due to the inherent defects of the imaging system or the influence of various interferences during the shooting process, the image is blurred, which brings troubles to people's applications. As the most discriminative part of the human body, the human face has received great attention from humans since ancient times. Portraits and self-portraits have always been popular subjects in paintings. After entering the era of mobile Internet, various face-based applications emerge in an en...

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): G06T5/00
CPCG06T2207/30201G06T5/73
Inventor王宇吴水清
OwnerHOHAI UNIV