An image quality assessment method based on multi-feature fusion in spatial domain and transform domain
A technology for image quality evaluation and multi-feature fusion, applied in image analysis, image enhancement, image data processing, etc., can solve the problems of not being able to describe the global image quality characteristics well, ignoring the human visual system, etc., and achieve model complexity The effect of good and predictive performance, balance between model complexity and predictive performance, good generalization ability and cross-database cross-running ability
Active Publication Date: 2021-05-25
XIAN UNIV OF TECH
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0007] At present, most full-reference image quality evaluation methods extract image features in the spatial domain, and most of the information in an image is concentrated in the low-frequency area, so it is not very good to only extract image features containing distortion information in the spatial domain. A good description of the global quality characteristics of the image ignores the characteristics of the human visual system that can process spatial and transform domain information at the same time, and there is still a lot of room for improvement in the accuracy of evaluating image quality
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0066] The present invention will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.
[0067] The present invention is an image quality evaluation method and technology based on multi-feature fusion of space domain and transform domain, such as figure 1 As shown, it is mainly divided into two parts: the establishment of image quality evaluation model and the prediction and evaluation of distorted image quality.
[0068] When establishing the image quality evaluation model, first extract the spatial gradient features, contrast sensitivity features, chromaticity features and visual saliency of all reference images and distorted images in the image database, after calculating the similarity, through multi-feature fusion technology respectively for each image The distorted image generates a 12-dimensional similarity feature vector, combined with the subjective MOS value, and uses the RF regression tool to train the quali...
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
Login to View More Abstract
An image quality evaluation method based on multi-feature fusion of space domain and transformation domain disclosed by the present invention, first performs color space conversion on all reference images and distorted images in the image database to prepare for subsequent channel-by-channel feature extraction; Extract the spatial gradient feature and contrast sensitivity feature of the image on the L channel, extract the chromaticity feature of the image on the two chrominance channels, and extract the visual saliency feature consistent with human visual perception in the global image. After the similarity calculation and pooling strategy, each distorted image can generate a 12-dimensional similarity feature vector; the similarity feature vector extracted from all distorted images in the image database and the subjective MOS value corresponding to the distorted image are input into the random forest RF training a regression model. The trained model can be used to accurately evaluate the quality of one or more distorted images to be evaluated.
Description
technical field [0001] The invention belongs to the technical field of image processing and image quality evaluation methods, and relates to an image quality evaluation method based on multi-feature fusion of space domain and transformation domain. Background technique [0002] With the vigorous development of digital media technology, multimedia communication information is flooding around us with its huge quantity and various forms of communication, enriching and changing our daily life, working methods and various production practices. Among them, digital images, as one of the main carriers of digital information, convey more realistic visual signals and bring people more diverse visual experiences, and have become an important part of our lives. [0003] During the process of image acquisition, transmission, processing, etc., the image quality will be degraded due to the influence of various factors. During acquisition, images captured by digital cameras, mobile phones,...
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
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06T7/00G06T7/90
CPCG06T7/0002G06T2207/30168G06T7/90
Inventor 郑元林于淼淼廖开阳唐梽森邓轩
Owner XIAN UNIV OF TECH