No-reference image quality evaluating method on basis of contourlet transform domain image energy features

Abstract

The invention discloses a no-reference image quality evaluating method on the basis of contourlet transform domain image energy features. The no-reference image quality evaluating method comprises the following steps of: (1) selecting ideal images, and then calculating all-level average energy after contourlet transform; (2) utilizing near linear relation among all-level energy features, and establishing a prediction model; (3) utilizing the prediction model and high-level average energy to predict low-level average energy; (4) carrying out preliminary measurement on the image quality according to the predicted average energy and the average energy calculated by distorted images; (5) carrying out energy compensation on the JPEG distorted type; and (6) constructing the no-reference image quality evaluating level according to the preliminary measurement of the image quality and the JPEG distortion. The no-reference image quality evaluating method does not need subjective score for training, is simple, can be applicable to multiple alteration types, is consistent to the subjective evaluation height and can be used for detecting the validity of an image-video processing method.

Description

technical field [0001] The invention relates to a non-reference image quality evaluation method based on contourlet (Contourlet) transform domain image features, and belongs to the technical field of image processing. Background technique [0002] Image quality is the main index to evaluate the performance of algorithms related to image and video processing, so the research on image quality evaluation methods is very important. There are two types of methods for evaluating image quality, subjective and objective. The end users of images are humans, so subjective methods are highly reliable, but subjective methods are time-consuming and laborious and are not easy to embed into automated systems. Therefore, the development of simple and effective Objective evaluation method is the focus of current image quality evaluation research. [0003] The objective evaluation methods are divided into three categories according to whether there are reference images, full reference qualit...

AI Technical Summary

Problems solved by technology

The author of the paper gives the experimental results of this method for JPEG2000 compression, JPEG compression, white noise, Gaussian blur and Fastfading five types of distortion, compared with other evaluation methods that can only be applied to one or two types of distortion. Improvement, but the effect of this method on JPEG and white noise distortion is very poor. In fact, it can only be applied to three types of distortion (JPEG2000 compression, Gaussian blur and Fastfading), so the development can be applied to more types of alteration. No reference image quality evaluation method is more valuable

Images