Image Quality Prediction Research Articles

Image Quality Assessment for Magnetic Resonance Imaging

Image quality assessment (IQA) algorithms aim to reproduce the human's perception of the image quality. The growing popularity of image enhancement, generation, and recovery models instigated the development of many methods to assess their performance. However, most IQA solutions are designed to predict image quality in the general domain, with the applicability to specific areas, such as medical imaging, remaining questionable. Moreover, the selection of these IQA metrics for a specific task typically involves intentionally induced distortions, such as manually added noise or artificial blurring; yet, the chosen metrics are then used to judge the output of real-life computer vision models. In this work, we aspire to fill these gaps by carrying out the most extensive IQA evaluation study for Magnetic Resonance Imaging (MRI) to date (14,700 subjective scores). We use outputs of neural network models trained to solve problems relevant to MRI, including image reconstruction in the scan acceleration, motion correction, and denoising. Our emphasis is on reflecting the radiologist's perception of the reconstructed images, gauging the most diagnostically influential criteria for the quality of MRI scans: signal-to-noise ratio, contrast-to-noise ratio, and the presence of artifacts. Seven trained radiologists assess these distorted images, with their verdicts then correlated with 35 different image quality metrics (full-reference, no-reference, and distribution-based metrics considered). The top performers -- DISTS, HaarPSI, VSI, and FID-VGG16 -- are found to be efficient across three proposed quality criteria, for all considered anatomies and the target tasks.

FQA-Net: an efficient neural network for blind image quality assessment

Guided by the neural network (NN), NN-based blind image quality assessment (BIQA) methods have improved image quality prediction accuracy. However, the training process of NN will consume a lot of time and training data. To solve the above problems, an efficient NN structure, which is called fast quality assessment network (FQA-Net), is proposed for BIQA. FQA-Net is a very simple NN, which mainly includes a convolution layer, a standard deviation measurement layer, and a regression layer. The convolution kernels in the convolution layer are a set of visual filters with characteristics of visual neurons, which are obtained by training on natural image samples. The output of the convolutional layer is a set of feature maps. Then the standard deviation of each feature map is calculated directly. Finally, a regression model is used for the mapping between the standard deviation values and the quality scores. FQA-Net not only reduces the number of parameters and the output dimensions in the training process, but also prevents network overfitting effectively. The results on seven image databases (i.e., LIVE, CSIQ, TID2013, LIVEMD, KADID-10K, LIVEC, and KonIQ-10K) show that FQA-Net has relatively low computational complexity and high computational speed and accuracy compared with other leading BIQA methods, especially when the number of training samples is very low. Moreover, the FQA-Net can provide stable prediction results across different distortion types.

A No-Reference Quality Assessment Method for Screen Content Images Based on Human Visual Perception Characteristics

The widespread application of screen content images (SCIs) has met the needs of remote display and online working. It is a topic that is challenging and worthwhile discussing in research on quality assessment for SCIs. However, existing methods focus on extracting artificial features to predict image quality, which are subjective and incomplete, or lack good interpretability. To overcome these problems, we propose an effective quality assessment method for SCIs based on human visual perceptual characteristics. The proposed method simulates the multi-channel working mechanism of the human visual system (HVS) through pyramid decomposition and the information extraction process of brains with the help of dictionary learning and sparse coding. The input SCIs are first decomposed at multiple scales, and then dictionary learning and sparse coding are applied to the images at each scale. Furthermore, the sparse representation results are analyzed from multiple perspectives. First, a pooling scheme about generalized Gaussian distribution and log-normal distribution is designed to describe the sparse coefficients with and without zero values, respectively. Then the sparse coefficients are used to characterize the energy characteristics. Additionally, the probability of each atom is calculated to describe the statistical property of SCIs. Since the above process only deals with brightness, color-related features are also added to make the model more general and robust. Experimental results on three public SCI databases show that the proposed method can achieve better performance than existing methods.

DMvLNet: deep multiview learning network for blindly accessing image quality

In image quality assessment (IQA), the quality of an image is essentially dependent on both semantic contents and image distortions. Conventional nonreference IQA (NR-IQA) methods either encode the semantic contents only or combine them in a straightforward manner without exploring their intrinsic correlation. A unified deep multiview learning network (DMvLNet) is proposed to combine both semantic contents and image distortions for NR-IQA. Using ResNet50 as the backbone, we design IQA-aware semantic features instead of directly using a recognition-specific CNN model for IQA, as in many existing methods. To this end, a high-level feature refining module built on deep layers is developed for generating refined high-level feature maps. Meanwhile, shallow layers are employed for low-level descriptions. Thus, both the refined high-level and low-level features are used for characterizing semantic contents and image distortions, respectively. Furthermore, the refined semantic features and distortion features are simultaneously integrated into DMvLNet by exploring their mutual correlation, leading to the projected multiview embedding passing through the subsequent transformer network for image quality prediction. Extensive experiments conducted on both our assembled large-scale dataset and seven public benchmark datasets demonstrate the superiority of DMvLNet over the state-of-the-art methods.

No-reference quality index of tone-mapped images based on authenticity, preservation, and scene expressiveness

As a future-proof technology, high dynamic range (HDR) has the superiority in scene reappearance. Tone mapping operator (TMO) renders the dynamic effect of HDR images in a standard dynamic range (SDR) display at the expense of some insignificant information. However, it is difficult for existing TMOs to consistently achieve high-quality conversion of all HDR images, and the loss of converted images is inevitable. Therefore, accurate prediction of tone-mapped image (TMI) quality is of considerable significance. In this paper, we propose the reappearance effect of the TMIs index (RETI). According to the characteristics of primeval HDR images, the quality of TMIs is considered from three important elements: authenticity, the preservation of energy and information, and scene expressiveness. The feature vectors originated from three elements are combined with subjective quality to train prediction model. Experiments are executed on tow mainstream TMI quality assessment databases and the results show that the proposed method has a good prediction ability and generalization ability in comparison to some state-of-the-art methods. Our MATLAB source code will be released at https://github.com/niannianyouyu/RETI.

HFFNet: hierarchical feature fusion network for blind binocular image quality prediction.

Compared with monocular images, scene discrepancies between the left- and right-view images impose additional challenges on visual quality predictions in binocular images. Herein, we propose a hierarchical feature fusion network (HFFNet) for blind binocular image quality prediction that handles scene discrepancies and uses multilevel fusion features from the left- and right-view images to reflect distortions in binocular images. Specifically, a feature extraction network based on MobileNetV2 is used to determine the feature layers from distorted binocular images; then, low-level binocular fusion features (or middle-level and high-level binocular fusion features) are obtained by fusing the left and right low-level monocular features (or middle-level and high-level monocular features) using the feature gate module; further, three feature enhancement modules are used to enrich the information of the extracted features at different levels. Finally, the total feature maps obtained from the high-, middle-, and low-level fusion features are applied to a three-input feature fusion module for feature merging. Thus, the proposed HFFNet provides better results, to the best of our knowledge, than existing methods on two benchmark datasets.

MVFStain: Multiple virtual functional stain histopathology images generation based on specific domain mapping.

To the best of our knowledge, artificial intelligence stain generation is an urgent requirement for histopathology images. Pathological examinations usually only utilize hematoxylin and eosin (H&E) regular staining to show histomorphological characteristics, but to accurately diagnose the disease, functional staining (such as oil red O and Ki67) are also required to provide important auxiliary information. However, the same tissue sectionis usually stained with one stain, and additional functional staining is not only time-consuming but also causes inevitable morphological misalignment due to manual manipulation. This brings difficulties to the development of artificial intelligence pathological image analysis tools. In this work, we propose a histopathology staining transfer framework to generate virtual functional staining images from H&E regular staining images. Compared with the framework that emphasizes generation diversity in the natural image field, we use KL loss and histo loss to align and separate style feature spaces in different domains to obtain domain-variant style features. The proposed multiple virtual functional stain (MVFStain) abstracts staining conversion to domain mapping and comprehensively utilizes multiple staining information. We evaluated the proposed method on four datasets (lung lesion, lung lobes, breast, and atherosclerotic lesion). The experiment involves the translation of H&E to nine other functional stains: CC10, Ki67, proSPC, HER2, PR, ER, oil red O, α-SMA, and macrophages. The major quantitative results are divided into image quality and positive signal prediction. MVFStain is close to or even surpasses one-to-one image translation on psnr and HTI image quality metrics. The best psnr reaches 26.1919, and HIT reaches 0.9430. We used mIOD to evaluate the optical density of positive signals, and CNR and gCNR to evaluate the lesion detectability. The results show that the mIOD of positive signals of virtual staining was slightly lower than the ground truth and close the lesion detectability of artificial staining. These results prove that the potential exists to develop a successful clinical alternative to artificial functional stains.

Wavelength-dependent Image Quality Predictions and Other Considerations for Site Selection and Instrument Design

Stellar image motion measured differentially with small telescopes can be converted to image quality predictions for large telescopes at the same site. To demonstrate this methodology, multi-wavelength seeing distributions are predicted for Las Campanas Observatory based on the Differential Image Motion Monitor (DIMM) results from the Giant Magellan Telescope site survey. The observed seeing distribution at 0.5 μm is used to compute a family of predicted seeing distributions from 0.32 to 4.7 μm. Tokovinin’s formula is used to transform the DIMM seeing distributions (based on Kolmogorov theory) to the large telescope image sizes expected in the von Karman theory. The latter framework includes an outer scale, or largest size, for the turbulent eddies above a telescope. The impact of seeing distributions on site selection and on telescope and instrument design are discussed. Other considerations in selecting an excellent site are also listed and explained.

Content-Variant Reference Image Quality Assessment via Knowledge Distillation

Generally, humans are more skilled at perceiving differences between high-quality (HQ) and low-quality (LQ) images than directly judging the quality of a single LQ image. This situation also applies to image quality assessment (IQA). Although recent no-reference (NR-IQA) methods have made great progress to predict image quality free from the reference image, they still have the potential to achieve better performance since HQ image information is not fully exploited. In contrast, full-reference (FR-IQA) methods tend to provide more reliable quality evaluation, but its practicability is affected by the requirement for pixel-level aligned reference images. To address this, we firstly propose the content-variant reference method via knowledge distillation (CVRKD-IQA). Specifically, we use non-aligned reference (NAR) images to introduce various prior distributions of high-quality images. The comparisons of distribution differences between HQ and LQ images can help our model better assess the image quality. Further, the knowledge distillation transfers more HQ-LQ distribution difference information from the FR-teacher to the NAR-student and stabilizing CVRKD-IQA performance. Moreover, to fully mine the local-global combined information, while achieving faster inference speed, our model directly processes multiple image patches from the input with the MLP-mixer. Cross-dataset experiments verify that our model can outperform all NAR/NR-IQA SOTAs, even reach comparable performance than FR-IQA methods on some occasions. Since the content-variant and non-aligned reference HQ images are easy to obtain, our model can support more IQA applications with its robustness to content variations. Our code is available: https://github.com/guanghaoyin/CVRKD-IQA.

Super Resolution Image Visual Quality Assessment Based on Feature Optimization.

Most existing no-referenced image quality assessment (NR-IQA) algorithms need to extract features first and then predict image quality. However, only a small number of features work in the model, and the rest will degrade the model performance. Consequently, an NR-IQA framework based on feature optimization is proposed to solve this problem and apply to the SR-IQA field. In this study, we designed a feature engineering method to solve this problem. Specifically, the features associate with the SR images were first collected and aggregated. Furthermore, several advanced feature selection algorithms were used to sort the feature sets according to their importance, and the importance matrix of features is obtained. Then, we examined the linear relationship between the number of features and Pearson linear correlation coefficient (PLCC) to determine the optimal number of features and the optimal feature selection algorithm, so as to obtain the optimal model. The results showed that the image quality scores predicted by the optimal model are in good agreement with the human subjective scores. Adopting the proposed feature optimization framework, we can effectively reduce the number of features in the model and obtain better performance. The experimental results indicated that SR image quality can be accurately predicted using only a small part of image features. In summary, we proposed a feature optimization framework to solve the current problem of irrelevant features in SR-IQA, and an SR image quality assessment model was proposed consequently.

Relating 18F-FDG image signal-to-noise ratio to time-of-flight noise-equivalent count rate in total-body PET using the uEXPLORER scanner

Objective. This work assessed the relationship between image signal-to-noise ratio (SNR) and total-body noise-equivalent count rate (NECR)—for both non-time-of-flight (TOF) NECR and TOF-NECR—in a long uniform water cylinder and 14 healthy human subjects using the uEXPLORER total-body PET/CT scanner. Approach. A TOF-NEC expression was modified for list-mode PET data, and both the non-TOF NECR and TOF-NECR were compared using datasets from a long uniform water cylinder and 14 human subjects scanned up to 12 h after radiotracer injection. Main results. The TOF-NECR for the uniform water cylinder was found to be linearly proportional to the TOF-reconstructed image SNR2 in the range of radioactivity concentrations studied, but not for non-TOF NECR as indicated by the reduced R 2 value. The results suggest that the use of TOF-NECR to estimate the count rate performance of TOF-enabled PET systems may be more appropriate for predicting the SNR of TOF-reconstructed images. Significance. Image quality in PET is commonly characterized by image SNR and, correspondingly, the NECR. While the use of NECR for predicting image quality in conventional PET systems is well-studied, the relationship between SNR and NECR has not been examined in detail in long axial field-of-view total-body PET systems, especially for human subjects. Furthermore, the current NEMA NU 2-2018 standard does not account for count rate performance gains due to TOF in the NECR evaluation. The relationship between image SNR and total-body NECR in long axial FOV PET was assessed for the first time using the uEXPLORER total-body PET/CT scanner.

Two Low-Level Feature Distributions Based No Reference Image Quality Assessment

No reference image quality assessment (NR IQA) aims to develop quantitative measures to automatically and accurately estimate perceptual image quality without any prior information about the reference image. In this paper, we introduce two low-level feature distributions (TLLFD) based method for NR IQA. Different from the deep learning method, the proposed method characterizes image quality with the distributions of low-level features, thus it has few parameters, simple model, high efficiency, and strong robustness. First, the texture change of distorted image is extracted by the weighted histogram of generalized local binary pattern. Second, the Weibull distribution of gradient is extracted to represent the structural change of the distorted image. Furthermore, support vector regression is adopted to model the complex nonlinear relationship between feature space and quality measure. Finally, numerical tests are performed on LIVE, CISQ, MICT, and TID2008 standard databases for five different distortion categories JPEG2000 (JP2K), JPEG, White Noise (WN), Gaussian Blur (GB), and Fast Fading (FF). The experimental results indicate that TLLFD method achieves superior performance and strong generalization for image quality prediction as compared to state-of-the-art full-reference, no reference, and even deep learning IQA methods.

A genetic programming-based convolutional neural network for image quality evaluations

Monitoring the perceptual quality of digital images is fundamentally important since digital image transmissions through the Internet continue to increase exponentially. Many automatic image quality evaluation (IQE) metrics have been developed based on image features correlated to image distortions; however, those metrics are only effective on particular image distortion types. In recent years, convolutional neural network (CNNs) have been developed for IQEs. These CNNs first capture image features from distorted images; image qualities are predicted based on the captured image features. Since the CNN weights are randomly initialized and are updated with respect to a loss function, image features which are strongly correlated to image quality are not guaranteed to be captured. In this paper, a hybrid deep neural network (DNN) is proposed by integrating image quality metrics to capture image features which are correlated to image quality; the approach guarantees that significant image features are included to predict image quality. Also, a tree-based classifier namely geometric semantic genetic programming is proposed to perform the overall predictions by incorporating CNN predictions and image features; the approach is simpler than the fully connected network but is able to model the nonlinear image qualities. The performance of the proposed hybrid DNN is evaluated by an image quality database with 3000 distorted images. The mean correlation achieved by the proposed hybrid DNN is 0.57 which is higher than the other tested methods. Experimental results with the t- test, F-test and Tueky’s range tests show that the proposed hybrid DNN achieves more accurate image predictions with a 99.9% confidence level, compared to the state-of-the-art IQE metics and the most recently developed CNN for IQEs.

Two birds with one stone: can [68Ga]Ga-DOTANOC PET/CT image quality be improved through BMI-adjusted injected activity without increasing acquisition times?

To assess how patients' dependent parameters may affect [68Ga]Ga-DOTANOC image quality and to propose a theoretical body mass index (BMI)-adjusted injected activity (IA) scheme, to improve imaging of high weight patients. Among patients prospectively enrolled (June-2019 and May-2020) in an Institutional Ethical Committee-approved electronic archive, we included those affected by primary gastro-entero-pancreatic (GEP) or lung neuroendocrine tumour and referred by our Institutional clinicians (excluding even minimal radiopharmaceutical extravasation, movement artefacts, renal insufficiency). All PET/CT images were acquired following EANM guidelines and rated for visual quality (1 = non-diagnostic, 2 = poor, 3 = moderate, 4 = good). Collected data included patient's body mass, height, BMI, age, IA (injected activity), IA/Kg (IAkg), IA/BMI (IABMI), liver SUVmean, liver SUVmax standard deviation, liver-signal-to-noise (LSNR), normalised_LSNR (LSNR_norm) and contrast-to-noise ratio (CNR) for positive scans and were compared to image rating (poor vs moderate/good). Overall, 77 patients were included. Rating concordance was high (agreement = 81.8%, Fleiss k score = 0.806). All patients' dependent parameters resulted significantly different between poor-rated and moderate/good-rated scans (IA: p = 0.006, IAkg: p =< 0.001, body weight: p =< 0.001, BMI: p =< 0.001, IABMI: p =< 0.001). Factors significantly associated with moderate/good rating were BMI (p =< 0.001), body weight (p =< 0.001), IABMI (p =< 0.001), IAkg (p = 0.001), IA (p = 0.003), LSNR_norm (p = 0.01). The BMI-based model presented the best predictive efficiency (81.82%). IABMI performance to differentiate moderate/good from poor rating resulted statistically significant (IA-AUC = 0.78; 95% CI: 0.68-0.89; cut-off value of 4.17 MBq*m2/kg, sensitivity = 81.1%, specificity = 66.7%). If BMI-adjusted IA (=4.17*BMI) would have been applied in this population, the median IA would have slightly inferior (-4.8%), despite a different IA in each patient. BMI resulted the best predictor of image quality. The proposed theoretical BMI-adjusted IA scheme (4.17*BMI) should yield images of better quality (especially in high-BMI patients) maintaining practical scanning times (3 min/bed).

Patient- and Examination-Related Predictors of 3D MRCP Image Quality in Children.

BACKGROUND. MRCP provides noninvasive imaging of the biliary tree and pancreatic duct. In our experience, MRCP image quality is commonly suboptimal in children. OBJECTIVE. The purpose of this study was to characterize the frequency of nondi-agnostic 3D fast spin-echo (FSE) MRCP acquisitions and determine predictors of nondi-agnostic MRCP image quality in children. METHODS. This retrospective study included 200 randomly selected pediatric patients (101 female and 99 male patients; mean age, 11.7 years) who underwent MRCP between January 1, 2019, and December 31, 2020. Patient- and examination-related variables were recorded. Three fellowship-trained pediatric radiologists independently reviewed 3D FSE MRCP acquisitions for diagnostic quality (diagnostic vs nondiagnostic) and overall image quality score on a scale from 1 to 5 (1 = worst image quality imaginable, 5 = best image quality imaginable). After computing interreader agreement, analyses used readers' most common diagnostic quality assessment and mean image quality score. Multivariable logistic regression and linear regression analyses were used to identify predictor variables of a diagnostic examination and higher image quality score. RESULTS. Interreader agreement for an MRCP acquisition being diagnostic quality, expressed as a kappa coefficient, was 0.53-0.71; interreader agreement for image quality score, expressed as an intraclass correlation coefficient, was 0.68-0.74. A total of 36 of 200 (18%) MRCP acquisitions were nondiagnostic; the mean image quality score was 3.5 ± 1.1 (SD). Multivariable predictors of a diagnostic MRCP acquisition included greater body mass index (OR = 1.11 [95% CI, 1.02-1.21]; p = .02), scanner field strength of 1.5 T (odds ratio [OR] = 2.87 [95% CI, 1.23-6.68]; p = .01), and presence of acute pancreatitis (OR = 4.91 [95% CI, 1.53-15.77]; p = .008). Multivariable predictors of a higher image quality score (β = 0.05-0.94) included older age (p = .01), imaging performed with patient under sedation or general anesthesia (p < .001), presence of biliary dilatation (p = .004), and inpatient status (p = .02). A lower image quality score was predicted by a scanner field strength of 3 T (β = -0.61; p < .001). A greater amount of time between the start of the MRI examination and the MRCP acquisition exhibited a nonsignificant association with a decrease in the image quality score (p = .06). CONCLUSION. Pediatric MRCP acquisitions are commonly nondiagnostic. Patient-specific and technical factors systematically impact MRCP image quality in children. CLINICAL IMPACT. Recognition of image quality predictors that are potentially modifiable and amendable to proactive intervention can guide efforts to optimize MRCP image quality in children.

Image Quality Predictor with Highly Efficient Fully Convolutional Neural Network

A highly efficient deep fully convolutional neural network (DFCN) for image quality assessment (IQA) is designed in this paper. The DFCN consists of two branches, one scoring local patches and the other estimating the weights of local patches to enhance quality prediction. Then, the DFCN outputs quality score of the whole image with aggregate weighted average pooling. There are no fully connected layers in the DFCN, resulting in far fewer parameters. In addition, the network model utilizes multiscale images as inputs to enrich the extracted distortion information. Furthermore, the parameters of the model are optimized in two steps to reduce the requirement for computing power and the risk of overfitting. The parameters of the shared layers and the quality module are optimized firstly, and then, the parameters of the weight module are optimized with the designed loss function. The extensive experimental results show that the proposed DFCN outperforms other competing IQA methods and has strong generalization ability.

Experimentally validated x-ray image simulations of 50 μm x-ray PIV tracer particles

We evaluate Beer–Lambert (BL) ray-tracing and Monte Carlo N-Particle (MCNP) photon tracking simulations for prediction and comparison of x-ray imaging system performance. These simulation tools can aid the methodical design of laboratory-scale x-ray particle image velocimetry (XPIV) experiments and tracer particles by predicting image quality. Particle image signal-to-noise ratio (SNR) is used as the metric of system performance. Simulated and experiment data of hollow, silver-coated, glass sphere tracer particles (AGSF-33) are compared. As predicted by the simulations, the AGSF-33 particles are visible with a SNR greater than unity in 100 ms exposure time images, demonstrating their potential as x-ray PIV or particle tracking velocimetry (XPTV) tracers. The BL approach predicts the image contrast, is computationally inexpensive, and enables the exploration of a vast parameter space for system design. MCNP simulations, on the other hand, predict experiment images slightly more accurately, but are more than an order of magnitude more computationally expensive than BL simulations. For most practical XPIV system design applications, the higher computational expense of MCNP is likely not justified by the modest accuracy improvement compared to BL.

Impact of left atrial enlargement on image quality in coronary computed tomography angiography

PurposeTo assess the associations between simple measurements of left atrial (LA) size and image quality of coronary computed tomography angiography (CCTA). Materials and methodsFour hundred and nineteen patients who underwent CCTA were retrospectively examined. Image quality was measured by coronary artery attenuation and contrast-to-noise ratio (CNR) of the proximal coronary artery (mean values of right coronary artery and left main trunk). LA transverse (LA-TRA) and anterior-posterior (LA-AP) diameter were measured on non-contrast CT images of the chest. The relationships of coronary attenuation and CNR with LA diameters were assessed by Pearson's correlation or Spearman's rank correlation coefficient and multivariate linear regression analysis. Receiver operating characteristic curves were used to assess the predictive value of LA diameters for image quality. ResultsBoth coronary artery attenuation value and CNR were independently correlated with LA-AP diameter (r = −0.38, for artery attenuation; r = −0.16, for CNR, both p < 0.001), and LA-TRA diameter (r = −0.2, p < 0.001, for artery attenuation; r = −0.11, p = 0.02, for CNR), respectively. With a cutoff value of 34 mm, the LA-AP diameter had a sensitivity of 85.2%, a specificity of 68.4% and an area under curve (AUC) of 0.77 for prediction of insufficient image quality (coronary artery attenuation less than 326 Hounsfield units). ConclusionCoronary artery attenuation and CNR in CCTA decreased with larger LA size.

No-reference image quality assessment for dehazed images

Traditional image quality prediction methods require the pristine image as a reference, such as structural similarity. However, it is difficult to provide a haze-free image as a reference when predicting the quality of the dehazed image. Therefore, it is necessary to use a no-reference image quality assessment (NR-IQA) method. In addition, most NR-IQA methods are based on known distortion type, using a large number of subjective opinion scores and images with the same distortion to train the model. We developed an innovative NR-IQA specifically for dehazed images without such prior knowledge. Since most images will undergo color distortion and blur after dehazing, it is proposed to combine color and sharpness for evaluation. The quality of the image is evaluated on the HSI color space, where the H and S channels are utilized to evaluate color, and the I channel to sharpness. Experimental results show that the performance of the proposed metric is better than other existing evaluation methods for dehazed images.

Study of Natural Scene Categories in Measurement of Perceived Image Quality

One challenge facing image quality assessment (IQA) is that current models designed or trained on the basis of exiting databases are intrinsically suboptimal and cannot deal with the real-world complexity and diversity of natural scenes. IQA models and databases are heavily skewed toward the visibility of distortions. It is critical to understand the wider determinants of perceived quality and use the new understanding to improve the predictive power of IQA models. Human behavioral categorization performance is powerful and essential for visual tasks. However, little is known about the impact of natural scene categories (SCs) on perceived image quality. We hypothesize that different classes of natural scenes influence image quality perception—how image quality is perceived is not only affected by the lower level image statistics and image structures shared between different categories but also by the semantic distinctions between these categories. In this article, we first design and conduct a fully controlled psychovisual experiment to verify our hypothesis. Then, we propose a computational framework that integrates the natural SC-specific component into image quality prediction. Research demonstrates the importance and plausibility of considering natural SCs in future IQA databases and models.