A Video Quality-Estimation Model for Streaming Media Services Based on Human Visual System

Streaming media services will become key services with advances in broadband IP networks. To provide a high-quality service for users, it is extremely important to design and manage the quality of experience (QoE) appropriately. To do this, it is necessary to develop an objective quality-assessment method that estimates subjective quality of streaming media. We propose a no-reference quality-estimation model for monitoring the video quality of streaming media services. The proposed model is based on video subjective quality assessment experiments and takes into account the spatial and temporal concealment of human visual system and the characteristic that human visual system is more sensitive to the parts with worse distortion of a video sequence. The experimental results indicate that Pearson Cross-Correlation was larger than 0.9, and the evaluation error was smaller than the statistical uncertainty of the value of subjective quality. Therefore, the proposed model could be applied to effective design, implementation, and management of streaming media services.