General-purpose image quality assessment based on distortion-aware decision fusion

Peng, Peng; Li, Ze-Nian

doi:10.1016/j.neucom.2013.08.046

Cited by 9 publications

(3 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Methods based on learning also exist: a support vector regression and parallel boosting measure is proposed in [6]. Two-stage framework that includes support-vector classification with following decision fusion of three metrics using k-nearest-neighbor was proposed in [10]. Despite training-based methods that use artificial neural networks and deep learning approaches could be very effective in solving of practical tasks there are interpretability and explainability issues, so we don't consider them in this work.…”

Section: Literature Reviewmentioning

confidence: 99%

Fine-Tuning of the Measure for Full Reference Image Quality Assessment

Gorokhovatskyi

Peredrii

2021

Lecture Notes in Computational Intelligence and Decision Making

View full text Add to dashboard Cite

In this paper, we proposed a new measure to solve the full reference image quality assessment problem. The core of the approach is known as peak signal-to-noise ratio improved with the estimation of local block-wise distortions, contrast, and saturation differences between test and referenced images. A measure that includes these values into a common quality score has been proposed. The proposed measure includes factors and thresholds which allow tuning the measure according to the specific features of the particular image dataset. The iterative numerical partial optimization algorithm for the fine-tuning of these factors and thresholds has been proposed, implemented, and tested as well as start optimization point selection has been described. The dependency of quality measure on parameters fine-tuning has been investigated. The usage of the proposed quality metric for the processing of TID2013 and CSIQ datasets as well as its computational complexity has been investigated. The results of modeling have been shown that it is possible to build the image quality measure in a fraction of a second preserving the average comparison quality in terms of the mean opinion score provided by humans.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Fine-Tuning of the Measure for Full Reference Image Quality Assessment

Gorokhovatskyi

Peredrii

2021

Lecture Notes in Computational Intelligence and Decision Making

View full text Add to dashboard Cite

show abstract

“…No-reference image quality assessment technique is proposed to monitor the image quality of processing systems and overcome the difficulty of reference images being absent [3]. This technique is deeply studied by many scholars [4][5][6][7][8]. The quality features are important factors that are extracted to represent the properties of image quality in the technique of no-reference image quality assessment [9].…”

Section: Introductionmentioning

confidence: 99%

No-reference image quality assessment based on quality patches in real time

Zhang

Yang

Huang

et al. 2018

J Image Video Proc.

View full text Add to dashboard Cite

Image quality assessment methods quantify the quality of an image that is highly correlated with human-perceived image quality. In this paper, the image quality is represented by the image patches that are selected by applying the spatial correlation of pixel-pairs. The quality image patches are selected from the database LIVE2, and quality feature detectors are learned by running the FastICA algorithm on the patches. Then being quality features, the independent component coefficients are found out from each quality image patch. A Hash lookup table is built by the binarization of the independent component coefficients of quality image patches, and the Hash table can be indexed by the binary code of the independent component coefficients of a tested image. The results of proposed approach were compared with results from other methods of image quality assessment and with the subjective image quality evaluated scores. And the experimental results expressed that the proposed metric method of no-reference image quality assessment could accurately measure the distorted degree of images. The Pearson linear correlation coefficient (PCC) achieves a high value 0.949 for the accuracy of evaluating results. The Spearman order correlation (SRC) achieves a high value 0.996 for the monotonicity of evaluating results. And the root mean square error (RMSE) is 5.917 for the subjective consistency of evaluating results. Extra evaluating aerial images, the proposed method obtained the result that the foggy weather led to the worst quality.

show abstract

“…Our method provides a new two-stage framework for existed methods to improve performances on their accuracy. Peng et al [37] also proposed an effective method for IQA by following the two-stage framework. However, their point that using the k-nearest-neighbor (KNN) to perform decision fusion of metrics can eliminate the limited ability of individual method across different distortion types was widely divergent.…”

Section: Introductionmentioning

confidence: 99%