Image database TID2013: Peculiarities, results and perspectives

Abstract

This paper describes a recently created image database, TID2013, intended for evaluation of full-reference visual quality assessment metrics. With respect to TID2008, the new database contains a larger number (3000) of test images obtained from 25 reference images, 24 types of distortions for each reference image, and 5 levels for each type of distortion. Motivations for introducing 7 new types of distortions and one additional level of distortions are given; examples of distorted images are presented. Mean Opinion Scores (MOS) for the new database have been collected by performing 985 subjective experiments with volunteers (observers) from five countries (Finland, France, Italy, Ukraine, and USA). The availability of MOS allows the use of the designed database as a fundamental tool for assessing the effectiveness of visual quality. Furthermore, existing visual quality metrics have been tested with the proposed database and the collected results have been analyzed using rank order correlation coefficients between MOS and considered metrics. These correlation indices have been obtained both considering the full set of distorted images and specific image subsets, for highlighting advantages and drawbacks of existing, state of the art, quality metrics. Approaches to thorough performance analysis for a given metric are presented to detect practical situations or distortion types for which this metric is not adequate enough to human perception. The created image database and the collected MOS values are freely available for downloading and utilization for scientific purposes

Acceleration of histogram‐based contrast enhancement via selective downsampling

“…Test images are collected from four standard databases, i.e., TID2013 [17], CSIQ [18], CCID2014 [10] and RGB-NIR [19]. TID2013 image dataset includes 25 reference images and their contrast-changed versions at Levels 1~5, which respectively correspond to small contrast decreasing/increasing, larger contrast decreasing/increasing and the largest contrast decreasing.…”

A Regression-Based Family of Measures for Full-Reference Image Quality Assessment

“…In the literature, some authors used different numbers of images from benchmarks for this purpose, e.g., 30% [13], [15], a one dataset [29], [24], [30], [18], [35], or even several datasets jointly, as in [20]. The following four IQA benchmarks were used: TID2013 [2], TID2008 [43], CSIQ [23], and LIVE [8], they are characterised in Table 1.…”

“…Contributions to rSIM1 2 The contribution of used measures seems to be equally distributed, except for PSNR in rSIM1 3 and IWSSIM in rSIM4 3 . Here, comparing rSIM1 2 to rSIM1 3 , and rSIM4 2 to rSIM4 3 , the second multimeasure is not influenced much by the presence of PSNR or IW-SSIM, respectively.…”

“…The conjunction of MAD with VIF seems to be interesting, both measures are less correlated (0.9632) with each other than with other measures, what may produce, together with their good single performance on LIVE dataset, the well-performing fusion measures. VSI is mostly used in fusion measures developed on images from TID2013, i.e., rSIM1 [2][3] . Here, VSI is the best single performing metric (see section 4), and that was the main reason for using this measure in the fusion.…”

“…Human evaluation of such images can be either inconvenient or expensive. Therefore, in order to measure image quality from the human perception point of view, many image quality assessment (IQA) measures have been developed [1], [2]. They are divided into three categories.…”