Super-resolution sparse reconstruction method based on discriminative canonical correlation

A super-resolution and sparse reconstruction technology, applied in image data processing, instrumentation, computing, etc., can solve problems such as lack of discrimination and insufficient use of class label information

Active Publication Date: 2016-11-09
刘姣姣
View PDF3 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

On the other hand, the CCA used by Huang's method is unsupervised, does not fully utilize the class label information of the training samples, and lacks certain discrimination

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
  • Super-resolution sparse reconstruction method based on discriminative canonical correlation
  • Super-resolution sparse reconstruction method based on discriminative canonical correlation
  • Super-resolution sparse reconstruction method based on discriminative canonical correlation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0041] In order to clarify the purpose, technical solutions and advantages of the present invention, the present invention will be further described in detail below in conjunction with specific embodiments and accompanying drawings.

[0042] refer to figure 1 , the specific implementation process of the present invention comprises the following steps:

[0043] (1) First, training samples for high-resolution images and corresponding low-resolution image samples Use principal component analysis (PCA) to extract the global features of high and low resolution face images, and obtain the projection vector P in the PCA subspace H ,P L , and its corresponding principal component X H 、X L . For the extracted PCA score vector X H 、X L , which is mapped to the relevant subspace of DCCA, and the corresponding projection vector C can be obtained by maximizing the criterion formula of DCCA H 、C L and score vector V H and V L .

[0044] Next, input a low-resolution test sampl...

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 provides a sparse face super-resolution reconstruction method based on discriminative canonical correlation. The extracted high and low resolution image features are projected to correlation subspaces by using supervised canonical correlation analysis, and neighborhood reconstruction is performed in the correlation subspaces by using sparse neighbor selection. The sample intra-class and inter-class correlation is fully considered, and supervision information is added so that the extracted features are enabled to be more discriminative, and the corresponding high-resolution images of test samples can be better reconstructed. Meanwhile, the sparse selection method is used in reconstruction neighbor selection, and the appropriate neighbors are self-adaptively selected for reconstruction according to the correlation of different training samples and the test samples so that the recovered global face images can be obtained. Compensation of detail information is performed by using the method and the idea of two-step method neighborhood reconstruction so that the high-resolution residual error images are reconstructed, and the global face images and the residual error images are added and thus the final high-resolution face images are obtained.

Description

technical field [0001] The invention relates to an image super-resolution reconstruction method. Specifically, it is a face image super-resolution sparse reconstruction method based on discriminant canonical correlation analysis, which can be applied to the fields of pattern recognition, data mining and image processing. Background technique [0002] Super-resolution (Super-resolution, SR) is an emerging technology that uses hardware or software to improve image resolution. By inputting one or more sets of low-resolution (Low Resolution, LR) images, they are processed by related algorithms. Finally, a set of high-resolution (High Resolution, HR) images are obtained. The use of super-resolution technology can achieve the purpose of improving the spatial resolution of images without changing the imaging system. , image registration, image quality evaluation and other research progress played a role in promoting. Super-resolution technology has broad application prospects, s...

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): G06T3/40
CPCG06T3/4053
Inventor 葛洪伟周梦璇李莉朱嘉钢
Owner 刘姣姣
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap