Bridge the Gap Between Full-Reference and No-Reference:A Totally Full-Reference Induced Blind Image Quality Assessment via Deep Neural Networks

Blind image quality assessment(BIQA)is of fundamental importance in low-level computer vision community.Increasing interest has been drawn in exploiting deep neural networks for BIQA.Despite of the notable success achieved,there is a broad consensus that training deep convolutional neural networks(DCNN)heavily relies on massive annotated data.Unfortunately,BIQA is typically a small sample problem,resulting the generalization ability of BIQA severely restricted.In order to improve the accuracy and generalization ability of BIQA metrics,this work proposed a totally opinion-unaware BIQA in which no subjective annotations are involved in the training stage.Multiple full-reference image quality assessment(FR-IQA)metrics are employed to label the distorted image as a substitution of subjective quality annotation.A deep neural network(DNN)is trained to blindly predict the multiple FR-IQA score in absence of corresponding pristine image.In the end,a selfsupervised FR-IQA score aggregator implemented by adversarial auto-encoder pools the predictions of multiple FR-IQA scores into the final quality predicting score.Even though none of subjective scores are involved in the training stage,experimental results indicate that our proposed full reference induced BIQA framework is as competitive as state-of-the-art BIQA metrics.

Blind Image Quality Assessment by Pairwise Ranking Image Series

Image quality assessment(IQA)is constantly innovating,but there are still three types of stickers that have not been resolved:the“content sticker”-limitation of training set,the“annotation sticker”-subjective instability in opinion scores and the“distortion sticker”-disordered distortion settings.In this paper,a No-Reference Image Quality Assessment(NR IQA)approach is proposed to deal with the problems.For“content sticker”,we introduce the idea of pairwise comparison and generate a largescale ranking set to pre-train the network;For“annotation sticker”,the absolute noise-containing subjective scores are transformed into ranking comparison results,and we design an indirect unsupervised regression based on EigenValue Decomposition(EVD);For“distortion sticker”,we propose a perception-based distortion classification method,which makes the distortion types clear and refined.Experiments have proved that our NR IQA approach Experiments show that the algorithm performs well and has good generalization ability.Furthermore,the proposed perception based distortion classification method would be able to provide insights on how the visual related studies may be developed and to broaden our understanding of human visual system.

Image quality assessment based on perceptual grouping

To further explore the human visual system（ HVS）,the perceptual grouping（ PG）, which has been proven to play an important role in the HVS, is adopted to design an effective image quality assessment（ IQA） model. Compared with the existing fixed-window-based models, the proposed one is an adaptive window-like model that introduces the perceptual grouping strategy into the IQA model. It works as follows： first,it preprocesses the images by clustering similar pixels into a group to the greatest extent; then the structural similarity is used to compute the similarity of the superpixels between reference and distorted images; finally, it integrates all the similarity of superpixels of an image to yield a quality score. Experimental results on three databases（ LIVE, IVC and MICT） showthat the proposed method yields good performance in terms of correlation with human judgments of visual quality.

No-reference image quality assessment based on AdaBoost_BP neural network in wavelet domain

Considering the relatively poor robustness of quality scores for different types of distortion and the lack of mechanism for determining distortion types, a no-reference image quality assessment(NR-IQA) method based on the Ada Boost BP neural network in the wavelet domain(WABNN) is proposed. A 36-dimensional image feature vector is constructed by extracting natural scene statistics(NSS) features and local information entropy features of the distorted image wavelet sub-band coefficients in three scales. The ABNN classifier is obtained by learning the relationship between image features and distortion types. The ABNN scorer is obtained by learning the relationship between image features and image quality scores. A series of contrast experiments are carried out in the laboratory of image and video engineering(LIVE) database and TID2013 database. Experimental results show the high accuracy of the distinguishing distortion type, the high consistency with subjective scores and the high robustness of the method for distorted images. Experiment results also show the independence of the database and the relatively high operation efficiency of this method.

Color Image Quality Assessment Based on Structural Similarity

It is well-known that classical quality measures,such as Mean Squared Error(MSE),Weighted Mean Squared Error(WMSE)or Peak Signal-to-Noise Ratio(PSNR),are not always corresponding with visual observations.Structural similarity based image quality assessment was proposed under the assumption that the Human Visual System(HVS)is highly adapted for extracting structural information from an image.While the demand on high color quality increases in the media industry,color loss will make the visual quality different.In this paper,we proposed an improved quality assessment(QA)method by adding color comparison into the structural similarity(SSIM)measurement system for evaluating color image quality.Then we divided the task of similarity measurement into four comparisons:luminance,contrast,structure,and color.Experimental results show that the predicted quality scores of the proposed method are more effective and consistent with visual quality than the classical methods using five different distortion types of color image sets.

NEW VISUAL PERCEPTUAL POOLING STRATEGY FOR IMAGE QUALITY ASSESSMENT

Most of Image Quality Assessment (IQA) metrics consist of two processes. In the first process, quality map of image is measured locally. In the second process, the last quality score is converted from the quality map by using the pooling strategy. The first process had been made effective and significant progresses, while the second process was always done in simple ways. In the second process of the pooling strategy, the optimal perceptual pooling weights should be determined and computed according to Human Visual System (HVS). Thus, a reliable spatial pooling mathematical model based on HVS is an important issue worthy of study. In this paper, a new Visual Perceptual Pooling Strategy (VPPS) for IQA is presented based on contrast sensitivity and luminance sensitivity of HVS. Experimental results with the LIVE database show that the visual perceptual weights, obtained by the proposed pooling strategy, can effectively and significantly improve the performances of the IQA metrics with Mean Structural SIMilarity (MSSIM) or Phase Quantization Code (PQC). It is confirmed that the proposed VPPS demonstrates promising results for improving the performances of existing IQA metrics.

SIMPLE QUALITY ASSESSMENT FOR BINARY IMAGES

Usually image assessment methods could be classified into two categories: subjective as-sessments and objective ones. The latter are judged by the correlation coefficient with subjective quality measurement MOS (Mean Opinion Score). This paper presents an objective quality assessment algorithm special for binary images. In the algorithm, noise energy is measured by Euclidean distance between noises and signals and the structural effects caused by noise are described by Euler number change. The assessment on image quality is calculated quantitatively in terms of PSNR (Peak Signal to Noise Ratio). Our experiments show that the results of the algorithm are highly correlative with subjective MOS and the algorithm is more simple and computational saving than traditional objective assessment methods.

Perceptual Quality Assessment of Omnidirectional Images:Subjective Experiment and Objective Model Evaluation

Virtual reality(VR) environment can provide immersive experience to viewers.Under the VR environment, providing a good quality of experience is extremely important.Therefore, in this paper, we present an image quality assessment(IQA) study on omnidirectional images. We first build an omnidirectional IQA(OIQA) database, including 16 source images with their corresponding 320 distorted images. We add four commonly encountered distortions. These distortions are JPEG compression, JPEG2000 compression, Gaussian blur, and Gaussian noise. Then we conduct a subjective quality evaluation study in the VR environment based on the OIQA database. Considering that visual attention is more important in VR environment, head and eye movement data are also tracked and collected during the quality rating experiments. The 16 raw and their corresponding distorted images,subjective quality assessment scores, and the head-orientation data and eye-gaze data together constitute the OIQA database. Based on the OIQA database, we test some state-of-the-art full-reference IQA(FR-IQA) measures on equirectangular format or cubic formatomnidirectional images. The results show that applying FR-IQA metrics on cubic format omnidirectional images could improve their performance. The performance of some FR-IQA metrics combining the saliency weight of three different types are also tested based on our database. Some new phenomena different from traditional IQA are observed.

Perceptual Gradient Similarity Deviation for Full Reference Image Quality Assessment

Perceptual image quality assessment(IQA)is one of the most indispensable yet challenging problems in image processing and computer vision.It is quite necessary to develop automatic and efficient approaches that can accurately predict perceptual image quality consistently with human subjective evaluation.To further improve the prediction accuracy for the distortion of color images,in this paper,we propose a novel effective and efficient IQA model,called perceptual gradient similarity deviation(PGSD).Based on the gradient magnitude similarity,we proposed a gradient direction selection method to automatically determine the pixel-wise perceptual gradient.The luminance and chrominance channels are both took into account to characterize the quality degradation caused by intensity and color distortions.Finally,a multi-scale strategy is utilized and pooled with different weights to incorporate image details at different resolutions.Experimental results on LIVE,CSIQ and TID2013 databases demonstrate the superior performances of the proposed algorithm.

A METHOD OF IMAGE QUALITY ASSESSMENT FOR COMPRESSIVE SAMPLING VIDEO TRANSMISSION

Based on compressive sampling transmission model, we demonstrate here a method of quality evaluation for the reconstruction images, which is promising for the transmission of unstructured signal with reduced dimension. By this method, the auxiliary information of the recovery image quality is obtained as a feedback to control number of measurements from compressive sampling video stream. Therefore, the number of measurements can be easily derived at the condition of the absence of information sparsity, and the recovery image quality is effectively improved. Theoretical and experimental results show that this algorithm can estimate the quality of images effectively and is in well consistency with the traditional objective evaluation algorithm.

No-Reference Image Quality Assessment Method Based on Visual Parameters

Recent studies on no-reference image quality assessment (NR-IQA) methods usually learn to evaluate the image quality by regressing from human subjective scores of the training samples. This study presented an NR-IQA method based on the basic image visual parameters without using human scored image databases in learning. We demonstrated that these features comprised the most basic characteristics for constructing an image and influencing the visual quality of an image. In this paper, the definitions, computational method, and relationships among these visual metrics were described. We subsequently proposed a no-reference assessment function, which was referred to as a visual parameter measurement index (VPMI), based on the integration of these visual metrics to assess image quality. It is established that the maximum of VPMI corresponds to the best quality of the color image. We verified this method using the popular assessment database—image quality assessment database (LIVE), and the results indicated that the proposed method matched better with the subjective assessment of human vision. Compared with other image quality assessment models, it is highly competitive. VPMI has low computational complexity, which makes it promising to implement in real-time image assessment systems.

Blind Image Quality Assessment Based on Hybrid Fuzzy-Genetic Technique

A new method for no-reference image quality assessment based on hybrid fuzzy-genetic technique is proposed. Noise variance and edge sharpness level of the restored image are two basic metrics for assessing the performance of the restoration algorithm, then a fuzzy if-then inference system is developed to combine the two metrics to get a final quality score, and the parameters of the fuzzy membership function are trained with genetic algorithms. Experiments results show that the image quality score correlates well with mean opinion score and the proposed approach is robust and effective.

No-Reference Quality Assessment of Enhanced Images

Image enhancement is a popular technique,which is widely used to improve the visual quality of images.While image enhancement has been extensively investigated,the relevant quality assessment of enhanced images remains an open problem,which may hinder further development of enhancement techniques.In this paper,a no-reference quality metric for digitally enhanced images is proposed.Three kinds of features are extracted for characterizing the quality of enhanced images,including non-structural information,sharpness and naturalness.Specifically,a total of 42 perceptual features are extracted and used to train a support vector regression(SVR) model.Finally,the trained SVR model is used for predicting the quality of enhanced images.The performance of the proposed method is evaluated on several enhancement-related databases,including a new enhanced image database built by the authors.The experimental results demonstrate the efficiency and advantage of the proposed metric.

Blind Image Quality Assessment Based on Wavelet Power Spectrum in Perceptual Domain

Blind image quality assessment(BIQA) can assess the perceptual quality of a distorted image without a prior knowledge of its reference image or distortion type. In this paper, a novel BIQA model is developed in wavelet domain. Considering the multi-resolution and band-passing characteristics of discrete wavelet transform(DWT), an improvement over the power spectrum is put forward, i.e., dubbed wavelet power spectrum(WPS)estimation. Then, the concept of directional WPS is applied to simplify the calculation. Moreover, a rotationally symmetric modulation transfer function(MTF) of human visual system(HVS) is integrated as a filter, which makes the metric to be consistent with the human vision perception and more discriminative. Experiments are conducted on the LIVE databases and three other databases, and the results show that the proposed metric is highly correlated with subjective evaluations and it competes well with other state-of-the-art metrics in terms of effectiveness and robustness.

No-Reference Stereo Image Quality Assessment Based on Transfer Learning

In order to apply the deep learning to the stereo image quality evaluation,two problems need to be solved:The first one is that we have a bit of training samples,another is how to input the dimensional image’s left view or right view.In this paper,we transfer the 2D image quality evaluation model to the stereo image quality evaluation,and this method solves the first problem;use the method of principal component analysis is used to fuse the left and right views into an input image in order to solve the second problem.At the same time,the input image is preprocessed by phase congruency transformation,which further improves the performance of the algorithm.The structure of the deep convolution neural network consists of four convolution layers and three maximum pooling layers and two fully connected layers.The experimental results on LIVE3D image database show that the prediction quality score of the model is in good agreement with the subjective evaluation value.

SAR Image Quality Assessment System Based on Human Visual Perception for Aircraft Electromagnetic Countermeasures

In electronic confrontation, Synthetic Aperture Radar (SAR) is vulnerable to different types of electronic jamming. The research on SAR jamming image quality assessment can provide the prerequisite for SAR jamming and anti-jamming technology, which is an urgent problem that researchers need to solve. Traditional SAR image quality assessment metrics analyze statistical error between the reference image and the jamming image only in the pixel domain; therefore, they cannot reflect the visual perceptual property of SAR jamming images effectively. In this demo, we develop a SAR image quality assessment system based on human visual perception for the application of aircraft electromagnetic countermeasures simulation platform.The internet of things and cloud computing techniques of big data are applied to our system. In the demonstration, we will present the assessment result interface of the SAR image quality assessment system.

Target acquisition performance in the presence of JPEG image compression

This paper presents an investigation on the effect of JPEG compression on the similarity between the target image and the background,where the similarity is further used to determine the degree of clutter in the image.Four new clutter metrics based on image quality assessment are introduced,among which the Haar wavelet-based perceptual similarity index,known as HaarPSI,provides the best target acquisition prediction results.It is shown that the similarity between the target and the background at the boundary between visually lossless and visually lossy compression does not change significantly compared to the case when an uncompressed image is used.In future work,through subjective tests,it is necessary to check whether this presence of compression at the threshold of just noticeable differences will affect the human target acquisition performance.Similarity values are compared with the results of subjective tests of the well-known target Search_2 database,where the degree of agreement between objective and subjective scores,measured through linear correlation,reached a value of 90%.

A NEW OBJECTIVE STEREOSCOPIC IMAGE ASSESSMENT MODEL BASED ON STEREOSCOPIC PERCEPTION

In order to establish a stereoscopic image quality assessment method which is consistent with human visual perception,we propose an objective stereoscopic image quality assessment method.It takes into account the strong correlation and high degree of structural between pixels of image.This method contains two models.One is the quality synthetic assessment of left-right view images,which is based on human visual characteristics,we use the Singular Value Decomposition(SVD)that can represent the degree of the distortion,and combine the qualities of left and right images by the characteristics of binocular superposition.The other model is stereoscopic perception quality assessment,due to strong stability of image’s singular value characteristics,we calculate the distance of the singular values and structural characteristic similarity of the absolute difference maps,and utilize the statistical value of the global error to evaluate stereoscopic perception.Finally,we combine two models to describe the stereoscopic image quality.Experimental results show that the correlation coefficients of the proposed assessment method and the human subjective perception are above 0.93,and the mean square errors are all less than 6.2,under JPEG,JP2K compression,Gaussian blurring,Gaussian white noise,H.264 coding distortion,and hybrid cross distortion.It indicates that the proposed stereoscopic objective method is consistent with human visual properties and also of availability.

Research on assessing compression quality taking into account the space-borne remote sensing images

According to the remote sensing image characteristics, a set oi optimized compression quahty assessment methods is proposed on the basis of generating simulative images. Firstly, a means is put forward that generates simulative images by scanning aerial films taking into account the space-borne remote sensing camera characteristics （including pixel resolution, histogram dynamic range and quantization）. In the course of compression quality assessment, the objective assessment considers images texture changes and mutual relationship between simulative images and decompressed ima- ges, while the synthesized estimation factor （SEF） is brought out innovatively for the first time. Subjective assessment adopts a display setup -- 0.5mrn/pixel, which considers human visual char- acteristic and mainstream monitor. The set of methods are applied in compression plan design of panchromatic camera loaded on ZY-1-02C satellite. Through systematic and comprehensive assess- ment, simulation results show that image compression quality with the compression ratio of d：l can meet the remote sensing application requirements.

Performance Validation and Analysis for Multi-Method Fusion Based Image Quality Metrics in A New Image Database

Considering that there is no single full reference image quality assessment method that could give the best performance in all situations, some multi-method fusion metrics were proposed. Machine learning techniques are often involved in such multi-method fusion metrics so that its output would be more consistent with human visual perceptions. On the other hand, the robustness and generalization ability of these multi-method fusion metrics are questioned because of the scarce of images with mean opinion scores. In order to comprehensively validate whether or not the generalization ability of such multi-method fusion IQA metrics are satisfying, we construct a new image database which contains up to 60 reference images. The newly built image database is then used to test the generalization ability of different multi-method fusion IQA metrics. Cross database validation experiment indicates that in our new image database, the performances of all the multi-method fusion IQA metrics have no statistical significant different with some single-method IQA metrics such as FSIM and MAD. In the end, a thorough analysis is given to explain why the performance of multi-method fusion IQA framework drop significantly in cross database validation.