Full-blind image quality evaluation method based on multi-dimensional visual feature cooperation under saliency modulation

A technology for image quality evaluation and visual features, which is applied in the field of image processing and can solve the problems of lack of fusion of subjective perception features of human eyes.

Active Publication Date: 2021-01-15
NORTHWEST UNIV
View PDF6 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008]The purpose of the present invention is to provide a kind of all-blind image quality evaluation method of multi-dimensional visual feature cooperation under saliency modulation, in order to solve the lack of subjective perception of human eyes in the prior art The problem of combining features

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
  • Full-blind image quality evaluation method based on multi-dimensional visual feature cooperation under saliency modulation
  • Full-blind image quality evaluation method based on multi-dimensional visual feature cooperation under saliency modulation
  • Full-blind image quality evaluation method based on multi-dimensional visual feature cooperation under saliency modulation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0043] A standard visual model construction method is disclosed in this embodiment, including the following steps:

[0044] Step a: Obtain multiple natural images, divide each natural image into image blocks to obtain multiple image blocks, and filter out image blocks that meet the sharpness threshold to obtain multiple high-sharpness image blocks;

[0045] Step b: For each high-sharpness image block obtained in step a, extract the natural scene statistical features in the spatial domain, the natural scene statistical features in the wavelet domain, the image structure features, the image color features and the image contrast features as image quality perception features, and obtain each The image quality perception features of each high-sharpness image block, and the image quality perception features of all high-sharpness image blocks are used as a feature vector matrix;

[0046] Step c: Fitting the eigenvector matrix obtained in step b with a multivariate Gaussian distributi...

Embodiment 2

[0108] In this embodiment, a blind image quality evaluation method based on multi-dimensional visual feature cooperation under saliency modulation is disclosed. On the basis of embodiment 1, the following technical features are also disclosed:

[0109] In step a, 125 high-quality natural images are selected, each image is divided into 96×96 image blocks, and M high-sharp image blocks are screened out.

[0110] In this embodiment, the three methods of NIQE, IL-NIQE, and IQA proposed by Zhang et al. are compared by experiments. The experimental results are shown in Table 1, where the Spearman rank correlation coefficient (SROCC) and the Pearson correlation coefficient (Pearson Correlation Coefficient, PLCC) are the evaluation indexes of the experiment, and the value is [0,1]. The higher the value, the The better the performance of the method.

[0111] Table 1 Comparison results between different methods

[0112]

[0113] It can be seen from the results in Table 1 that the b...

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 belongs to the technical field of image processing, and discloses a full-blind image quality evaluation method based on multi-dimensional visual feature cooperation under saliency modulation, and the method comprises the steps: obtaining a distorted image block of a to-be-detected distorted image, and extracting an image quality perception feature; taking the image quality perceptionfeatures of all distorted image blocks as a to-be-measured feature vector matrix; fitting the obtained feature vector matrix to be measured based on visual saliency to obtain a visual model to be measured; and finally, calculating the Mahalanobis distance between the visual model to be measured and the standard visual model to obtain the objective quality score of the distorted image to be measured. According to the method, a feature descriptor used for expressing image contrast distortion and hue distortion is constructed in combination with human vision primary perception features, and high-order natural scene statistical features, image structure features and color features of the image are combined, so that image distortion is expressed more comprehensively.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to an image quality evaluation method, in particular to an all-blind image quality evaluation method based on multi-dimensional visual feature cooperation under saliency modulation. Background technique [0002] Blind Image Quality Assessment (BIQA) is a research task in the field of image processing aimed at designing a computational model that does not rely on any prior knowledge and can automatically evaluate image quality. The performance provides an important basis for the research in other fields of image processing. [0003] The existing blind image quality assessment methods include two categories: one is the Opinion-aware (OA) BIQA method. This type of method first needs to establish an image dataset, which contains multiple distorted images, and each image has a corresponding human eye average subjective score (Mean Opinion Score, MOS) or average subjective score d...

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): G06T7/00G06T7/11G06T7/90
CPCG06T7/0004G06T7/11G06T7/90G06T2207/10024G06T2207/30168
Inventor 张敏侯文静许筱敏张蕾冯筠吕毅
Owner NORTHWEST UNIV
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