Perceptual visual quality metrics: A survey

By Weisi Lin, C.-C. Jay Kuo. Journal of Visual Communication and Image Representation, 2011

968 citations recorded by the source corpus.

Visual quality evaluation has numerous uses in practice, and also plays a central role in shaping many visual processing algorithms and systems, as well as their implementation, optimization and testing. In this paper, we give a systematic, comprehensive and up-to-date review of perceptual visual quality metrics (PVQMs) to predict picture quality according to human perception. Several frequently used computational modules (building blocks of PVQMs) are discussed. These include signal decomposition, just-noticeable distortion, visual attention, and common feature and artifact detection. Afterwards, different types of existing PVQMs are presented, and further discussion is given toward feature pooling, viewing condition, computer-generated signal and visual attention. Six often-used image metrics(namely SSIM, VSNR, IFC, VIF, MSVD and PSNR) are also compared with seven public image databases (totally 3832 test images). We highlight the most significant research work for each topic and provide the links to the extensive relevant literature.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/76d82dd336337cce077715282c34ff6ec1a44d2d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

From QoS to QoE: A Tutorial on Video Quality Assessment (2015, methodology): “For example, [32], [69] mainly follow the psychophysical/engineering approach classiﬁcation method; [31], [70]–[72] mainly adopt the reference-based clas-siﬁcation method, and [47] adopts a combination of the two.”
No-Reference Quality Metric of Contrast-Distorted Images Based on Information Maximization (2017, background): “This promotes the development of objective image quality assessment (IQA) metrics, which predict the visual quality using mathematical models to simulate subjective opinion scores [1].”
The Analysis of Image Contrast: From Quality Assessment to Automatic Enhancement (2016, background): “databases [4], [5], existing objective IQA techniques are mainly devoted to compression, transmission error, noise, and blurring artifacts.”
Perceptual Quality Assessment of Smartphone Photography (2020, background): “In the past decades, there have been a significant number of studies on both directions [1, 12, 17], most of which focus on synthetic distortions, with the assumption that the original undistorted images exist and can be used as reference [37].”
Seven Challenges in Image Quality Assessment: Past, Present, and Future Research (2013, background): “Both Sheikh et al. [119] and Lin and Kuo [35] have provided analyses of the performances of various IQA algorithms on some of these databases.”
MCL-JCV: A JND-based H.264/AVC video quality assessment dataset (2016, methodology): “However, the mean-squared-errors (MSE) distortion measure used in these standards has been criticized for not being well correlated with human visual experience.”
Color image database TID2013: Peculiarities and preliminary results (2013, background): “In modern applications of digital image processing, visual quality is of a prime importance [1-3].”
Image Quality Assessment Based on Gradient Similarity (2012, neutral): “Aimed at accurately and automatically evaluating the image quality in a manner that agrees with subjective human judgments (HVS-oriented), regardless of the type of distortion corrupting the image, the content of the image, or the strength of the distortion, substantial research effort has been dire”
No-Reference Quality Assessment of Contrast-Distorted Images Based on Natural Scene Statistics (2015, background): “To overcome these limitations, many objective IQA metrics have been proposed during the past decades [1] to predict human perception of image quality.”
PieAPP: Perceptual Image-Error Assessment Through Pairwise Preference (2018, background): “Because of its importance, this problem, also known as full-reference image-quality assessment (FR-IQA) [58], has received significant research attention in the past few decades [5, 11, 23, 29, 30, 32].”
Blind Quality Assessment Based on Pseudo-Reference Image (2018, background): “During the past decades, dozens of IQA models have been proposed [2]–[4].”
Deep Neural Networks for No-Reference and Full-Reference Image Quality Assessment (2016, background): “This led to the development of a whole zoo of image quality metrics that strive for a better agreement with the image quality as perceived by humans [7].”
Mean opinion score (MOS) revisited: methods and applications, limitations and alternatives (2016, background): “Large numbers of subjective MOS data [94] and objective quality metrics [37-39] have been made available.”
TOPIQ: A Top-Down Approach From Semantics to Distortions for Image Quality Assessment (2023, methodology): “According to comprehensive survey of IQA [2], [3], [4], most IQA techniques can be categorized as Full-Reference (FR) IQA or No-Reference (NR) IQA based on the requirement for pristine reference images.”
Image database TID 2013 : Peculiarities , results and perspectives (2016, background): “Many visual quality metrics have been exploited in applications such as image and video lossy compression, image denoising, quality control of image printing and scanning, remote sensing, watermarking [2] [3] [4] [5] [6] [7] [8], etc.”