Video Quality Assessment Model Based On Space And Time Domain

With the rapid development of IP networks,3G networks and LTE, digital video communications has become an important area, and video services will become the support of the communications business. In the development process of video business, monitoring just on video protocol and video transmission has been unable to meet the requirements for video quality, and video service quality has become a new requirement point.Traditional video quality assessment models often focus only on the absolute error rate in the video, which can not reflect the impact of the human eye on the assessment results, and often make the assessment results deviate from subjective feelings. So the purpose of this study is that focusing on a non-reference video quality assessment based on space-time domain from users'perspective, objectively reflecting the quality of video services, and providing a strong guarantee for network video services.Through analysis and comparison of different video quality assessment methods, this paper propose a non-reference objective video quality assessment model based on space domain and time domain in order to meet real-time monitoring of network video business. The main contents of this thesis is that:based on spatial indicator blocking and temporal indicator dynamic, adding image edge features, considering the characteristics of human visual system, and finally developing a non-reference evaluation algorithm based on space-time domain,the results of which is more consistent with human visual characteristics. Experimental results show that:video quality assessment results of the method have good consistency with the subjective evaluation results. In order to make the video quality assessment model using for real-time network monitoring, this paper implements this model as video quality analysis components, describing in detail the structure of this component design, the methods of implementation, and the application in the video monitoring of IP network,3G network and other actual video networks.