Algoritmo VQM-SSIM para Avaliação da Degradação da Qualidade de Conteúdo Multimídia Transmitido por Dispositivo Android

Marcia dos Santos de Sa; Jucicarla Pires Barbosa; Ewerton M. Barbosa; Luiz Eduardo S. de Araújo; Walter Seiffert Simões

doi:10.20906/SBAI-SBSE-2023/3773

Marcia dos Santos de Sa ICOMP - Instituto de Computação - Universidade Federal do Amazonas - UFAM – 69.067-005 – Manaus – AM & CESAR - Centro de Estudos e Sistemas Avançados de Recife - Manaus - AM
Jucicarla Pires Barbosa ICOMP - Instituto de Computação - Universidade Federal do Amazonas - UFAM – 69.067-005 – Manaus – AM
Ewerton M. Barbosa ICOMP - Instituto de Computação - Universidade Federal do Amazonas - UFAM – 69.067-005 – Manaus – AM
Luiz Eduardo S. de Araújo ICOMP - Instituto de Computação - Universidade Federal do Amazonas - UFAM – 69.067-005 – Manaus – AM
Walter Seiffert Simões ICOMP - Instituto de Computação - Universidade Federal do Amazonas - UFAM – 69.067-005 – Manaus – AM & CESAR - Centro de Estudos e Sistemas Avançados de Recife - Manaus - AM

DOI: https://doi.org/10.20906/SBAI-SBSE-2023/3773

Keywords: Video Quality, SSIM, PSNR, Linear Regression, Chaquopy

Abstract

Currently, streaming video content is subjectively evaluated by smartphone manufacturers to indicate potential flaws perceived in receiving the content. However, many flaws are not noticed and can generate unpleasant effects for the end user if they are not worked on. This paper proposes to use Artificial Intelligence (AI) techniques to align the user’s visual demands with an objective evaluation. The applied methodology consists of combining the Video Quality Metric (VQM) and Structural Similarity Index (SSIM) algorithms by means of AI techniques, in order to provide an indication of the degradation of the multimedia content without the interference of possible transmission noise in the data network. The results show a level of accuracy in the similarity analysis of static multimedia structure at 99.42% and 99.10% for dynamic media, consumed in test time. The major contribution of this work is in the computing society, which will be able to apply an objective tool in the evaluation of multimedia content.