Visual distortion gauge based on discrimination of noticeable contrast changes

Abstract

This paper presents a method to discriminate pixel differences according to their impact toward perceived visual quality. Noticeable local contrast changes are formulated firstly since contrast is the basic sensory feature in the human visual system (HVS) perception. The analysis aims at quantifying the actual impact of such changes (further divided into increases and decreases on edges) in different signal contexts. An associated full-reference distortion metric proposed next provides better match with the HVS viewing. Experiments have used two independent visual data sets and the related subjective viewing results, and demonstrated the performance improvement of the proposed metric over the relevant existing ones with various video/images and under diversified test conditions. The proposed metric is particularly effective to visual signal with blurring and luminance fluctuations as the major artifacts, and brings about the fundamental improvement when sharpened image edges are involved.Index Terms-Edge contrast increase, human visual system (HVS), just noticeable difference (JND), perceptual visual quality, subjective quality ratings.

A human visual system-driven image segmentation algorithm

“…JND refers to the minimum visibility threshold above which visual contents can be distinguished. The JND model plays an important role in perceptual image and video processing, and has been successfully used in measuring the difference or distortion of the image contents [40] [41]. In this paper, we use the spatial JND model, i.e., pixel-wise JND, presented in [42], which is defined as a nonlinear additive model, …”

Perceptual distortion modeling for side-by-side 3D video delivery

“…The component related to intensity contrast is based on the higher sensitivity of the HVS to luminance variations with respect to the environment of a stimulus, than to the absolute luminance of the stimulus, as stated by the Weber-Fechner law. In fact, a piecewise approximation of the sensitivity threshold was used [6], Moreover, the temporal masking factor is modeled as a function of the luminance difference between corresponding MB of contiguous frames [5]. Finally, the threshold related to the local activity of the environment of the analyzed MB is computed from the difference between the maximum and the minimum mean luminance of the neighboring MBs [6].…”

“…In fact, a piecewise approximation of the sensitivity threshold was used [6], Moreover, the temporal masking factor is modeled as a function of the luminance difference between corresponding MB of contiguous frames [5]. Finally, the threshold related to the local activity of the environment of the analyzed MB is computed from the difference between the maximum and the minimum mean luminance of the neighboring MBs [6]. In addition, to reflect the effects related to the fusion of both stereo views, corresponding MBs in each stereo view are used in the computation of the luminance and texture thresholds.…”

“…These features are based on a Just Noticeable Distortion (HMD) model, since they can provide information about the maximum distortion that the signal can suffer without being perceived [5]. These models have been extensively used in the field of audio and video compression and also applied to quality metrics, thanks to their direct relation to the perceptual behavior of the human perception [6],…”

Modeling Blocking Visual Sensitivity Profile

“…And these perceptual-plausible methods show excellent performance in many applications. However, much attention of engineers in this field goes into estimating the masking effects on a whole-image level and they often compute perceptual models for every pixel in the image [3][4][5]. Rather than that, in this paper, we emphasize on modeling the masking effects on blocking artifacts and estimate their visual sensitivity.…”