Carlos Danilo Miranda Regis

Objective estimation of 3D video quality: A disparity-based weighting strategy

This paper presents new results on the objective evaluation of stereoscopic video quality. A balance between the scores provided by classical objective algorithms and the disparity information of the reference video is found. The performance of the proposed technique was verified using statistical metrics (correlation coefficients and root mean square error) that compare the scores provided by the objective algorithms and the subjective results provided by the NAMA3DS1-COSPAD1 stereoscopic video quality database. The obtained results suggest a significant improvement on the performance of the objective algorithms for the proposed technique.

Effect of visual attention areas on the objective video quality assessment

The objective assessment of video quality is an important area of research for the validation of video systems, and new algorithms are always in demand. Experiments indicate that the inclusion of characteristics of the human visual system into the objective metric, such as visual attention, improves the evaluation. This paper presents a new approach to evaluate the video quality based on an estimate of the local visual attention, using the structural similarity index. This approach is compared with three classical objective metrics: PSNR, SSIM and MS-SSIM, by means of the Pearson linear and Spearman rank-order correlation coefficients. The videos are submitted to several types of distortions, such as blocking, blurring and salt & pepper noise. The results indicate that the proposed approach presents good correlation with the subjective scores, for videos that contain blurring and blocking degradations.

Performance of the objective video quality metrics with perceptual weighting considering first and second order differential operators

Objective video quality assessment is important to validate the performance of applications in video communications systems. The research on visual attention and gradient-inspired metrics improves the performance of the video quality evaluation. This paper investigates the use of first and second order differential operators as visual attention estimators. The edge information, computed by differential operators, is an important factor of visual attention. The performance of the proposed models are compared with the objective metrics, PSNR, SSIM and GSSIM, by means of the Pearson correlation coefficient, considering the blurring, blocking, salt and pepper and Gaussian noise distortions. The results suggest that for blurring, blocking and salt and pepper, the combination of objective metrics with first order operators provides a significant improvement, especially for the Sobel operator.

Temporal analysis and perceptual weighting for objective video quality measurement

Video quality measurement is an important factor to establish the quality of video communication and processing systems, such as broadcasting and mobile television, since they need to monitor broadcasting quality to compute optimal parameters that ensure a satisfactory quality level. The objective video quality assessment is a quick and low cost alternative when compared with the subjective evaluations. This paper presents a novel objective video quality algorithm that combines temporal analysis and the perceptual weighting technique. The proposed algorithm was validated using the correlation between objective measurements and the subjective scores provided by the LIVE Video Quality Database, considering the following conditions: H.264 and MPEG-2 encoding and transmission of H.264 bit-streams over IP and wireless networks. The results demonstrate that the proposed algorithm is a competitive alternative when compared with classical objective algorithms, such as MOVIE.

Disparity weighting applied to full-reference and no reference stereoscopic image quality assessment

This paper presents an evaluation of performance of the disparity weighting technique when it is applied to stereoscopic image quality assessment in full reference and no-reference scenarios. The correlation coefficients between the subjective scores in the LIVE 3D Database and the predicted scores indicate a significant increase in the performance.

Design of Objective Video Quality Metrics Using Spatial and Temporal Informations

Digital videos are subject to several types of distortions, which may occur during the processes of acquisition, processing, compression, storage and transmission of videos, resulting in loss of visual quality. Video quality assessment is important to set system parameters and propose the reduction of degradation, to improve the systems quality. The most appropriate way to assess the visual quality of a video sequence is subjective assessment. This study proposes new metrics for assessment of videos. The proposed metrics that uses spatial and temporal information to evaluate videos are more closely to the subjective assessment. For the evaluation of videos the BD-TPW-SSIM and TP-VQI metric obtained the best result for wireless degradation. For MPEG-2 encoded videos, T-PW-SSIM e T-MOVIE was the best, and IP systems the metric with the best result was the TPW-SSIM. Finally, the metric T-VimSSIM was the best for H.264 encoding.

MUNIQUE: Multiview no-reference image quality evaluation

This paper presents a novel no-reference objective algorithm for stereoscopic image quality assessment, called MUNIQUE, which is based on the estimation of both two-dimensional and stereoscopic features of images, namely local estimations of blockiness and blurriness and the disparity weighting technique. Applications of stereoscopic image and video quality assessment in surveillance systems are discussed. Simulation results using LIVE 3D Image Quality Database Phase I, which includes Gaussian blur and fast fading degraded images, are presented and a comparison of performance of MUNIQUE with several state of the art algorithms is made. Correlation coefficients between subjective and predicted scores indicate a superior performance of the proposed algorithm, when it is compared with others no-reference algorithms. An implementation of the proposed algorithm coded in C# programming language is publicly available at: https://sites.google.com/site/jvmircas/home/munique.

On the performance of disparity-based weighting technique applied to 3D image quality assessment

This papers describes the performance of objective algorithms when they are applied for stereosocopic image quality assessment. Specifically, the authors ascertained the performance of the objective algorithms after the combination with an weighting by means of the disparity. The IRCCyN-IVC 3D Images Quality database was used to verify the performance of the objective algorithms. Indeed, the obtained results suggest a significative enhance of the prediction capacity of the objective algorithms when they were combined with the disparity information contained in the original image signal.

ImQET: Objective Stereoscopic Image Quality Evaluation Tool

This paper describes an application for full-reference stereoscopic image quality assessment called ImQET. The application was developed using Mono framework and C# programming language. ImQET is independent of platform and provides a friendly Graphical User Interface (GUI). The stereoscopic image signals used in the application are based in a two-view model. ImQET has objective image quality algorithms such as PSNR, SSIM, and PW-SSIM and also incorporates a recently published technique for stereoscopic image quality assessment called Disparity Weighting (DW). Numerical results corresponding to the performance of the objective measurements, which were obtained using the proposed application, are presented. ImQET can be used by academia and industry for standardization and development of objective algorithms and evaluation of impairments in stereoscopic image signals caused by processing techniques.

SQUALES: A Qt-based application for full-reference objective stereoscopic video quality measurement

This paper describes a Qt-based C++ application for full-reference stereoscopic video quality assessment. The application is platform independent and provides a friendly Graphical User Interface (GUI). The stereoscopic video signals used in the application are based in two-view model, such as the H.264/AVC standard with Multiview Video Coding (MVC), and several spatial resolutions are available. The application provides objective video quality algorithms, such as PSNR, SSIM and PW-SSIM and incorporates a recently published technique for stereoscopic video quality assessment, called disparity weighting (DW), DPSNR, DSSIM and DPW-SSIM. Numerical results corresponding to the performance of the objective measurements acquired using the proposed application are presented. The ultimate goal is that the application will contribute, as an open-source tool to be used by academia and industry, for standardization and development of objective algorithms and evaluation of impairments in stereoscopic video signals caused by some processing techniques.