Unlock instant, AI-driven research and patent intelligence for your innovation.

Accurate no-reference image quality evaluation method based on distortion identification

A technology of reference image and quality evaluation, applied in image enhancement, image analysis, image data processing, etc.

Active Publication Date: 2020-09-01
HANGZHOU DIANZI UNIV
View PDF11 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Solve the evaluation problem of multi-distortion scenes through a general quality evaluation model based on image distortion types

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
  • Accurate no-reference image quality evaluation method based on distortion identification
  • Accurate no-reference image quality evaluation method based on distortion identification
  • Accurate no-reference image quality evaluation method based on distortion identification

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0087] The content of the present invention will be further described below in conjunction with the accompanying drawings.

[0088]In general, the difficulty of no-reference image quality assessment lies in its blindness and low efficiency due to insufficient grasp of image distortion information. In order to solve this problem, the present invention proposes a new evaluation strategy, which is divided into two steps of distortion identification and targeted quality evaluation. In the first step, we train a classifier using the Inception-Resnet-v2 neural network to classify possible distortions in images into the four most common types of distortion: Gaussian noise, Gaussian blur, jpeg compression, jpeg2000 compression. In the second step, after determining the type of image distortion, we design a specific method to quantify the degree of image distortion, so that the quality of the image can be evaluated more accurately. Our preliminary experiments on LIVE, TID2013, CSIQ an...

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 an accurate no-reference image quality evaluation method based on distortion identification. According to the method, distortion information of a to-be-evaluated image is accurately classified, then targeted evaluation is carried out according to specific type features, in the aspect of classification work, an Inception-Resne-v2 neural network is used for training a classifier, distortion classification is carried out, and a type label is output; and finally, corresponding evaluation work is carried out according to the type label output in the step (2). According to the method, a new open type evaluation strategy is designed, a model which is classified firstly and then evaluated is designed by simulating subjective evaluation logic of people, the problem that common features in a universal model are difficult to design is indirectly solved, and the evaluation precision is higher than that of other methods of the same type.

Description

technical field [0001] The invention relates to the field of image quality evaluation, and specifically proposes a no-reference quality evaluation method based on accurate distortion recognition. Background technique [0002] Digital images are ubiquitous in our lives, and it is mainly transmitted through digital devices and applications in our lives, such as HDTV, video chat or Internet video streaming, etc. But in the process of transmission, the quality of the image basically has a certain loss. Constraints such as exposure time, light sensitivity, aperture, and lenses all affect image quality and ultimately lead to poorer perceived visual quality. In general practical scenarios, the visual quality of images is mainly evaluated by humans, and this subjective quality evaluation method is also considered to be the most accurate and reliable evaluation method. However, collecting subjective scores in experiments is quite laborious and expensive, so how to design an objecti...

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): G06T7/00G06K9/62G06N3/04G06N3/08
CPCG06T7/0002G06T2207/10016G06T2207/10024G06T2207/20024G06T2207/20064G06T2207/20081G06T2207/20084G06T2207/30168G06N3/08G06N3/045G06F18/241
Inventor 颜成钢滕统孙垚棋张继勇张勇东
Owner HANGZHOU DIANZI UNIV