Fractal Analysis of Radon Coefficients for No-Reference Video Quality Assessment (NR-VQA)

Abstract

Forecast video quality in the absence of a reference video is a difficult task. This paper proposes a novel method for evaluating the quality of videos using the Radon transform. We propose fractal analysis of Radon coefficients to determine the video quality in the absence of reference data. Fractal analysis is a mathematical technique for characterizing the properties of objects that have an irregular or complex structure. It is used to extract the structural changes that occur in video frames as a result of various distortions. Additionally, a support vector regression model is used to predict the quality score of the video. Three widely used and publicly available video quality databases are used to validate the proposed NR-VQA model. In terms of quality prediction performance, the proposed model outperforms the majority of existing state-of-the-art methods.