Subjective and objective quality assessment of MPEG-2, H.264 and H.265 videos

Abstract

With the proliferation of video based applications, the interest for video quality assessment (VQA) has been rising. Subjective VQA experiments are expensive and time consuming. Alternatives are objective VQA algorithms. Different video databases are created and used for objective VQA algorithms development and performance evaluation. This paper presents a new video quality database called FERIT-RTRK database, which contains videos degraded according to three different video compression norms: MPEG-2, H.264 and H.265. The paper also reports information about related conducted subjective VQA experiment and compares data obtained by the subjective VQA experiment to scores calculated by four widely used objective VQA algorithms. The new database can be used for performance evaluation of state-of-the-art objective VQA algorithms, since it contains videos degraded by the compression norms which are currently widely used at market. The database and all relevant information are publicly available at http://www.rtrk.com/other/VideoDBReadme.html.