User Experience Oriented Quality Evaluation For Video Streaming Service

The development of video service promotes the user demand for high quality real-time video streaming service. The performance assessing has become more and more important. The traditional multimedia quality evaluation indicators are based on quality of service (QoS). However, the network QoS cannot reflect increase users’ satisfaction degree effectively for video services, because QoS cannot effectively reflect users’actual demand and subjective experience. How to assess the end-to-end user-oriented video service quality more effectively become an urgent problematical to solve. Then, quality of experience (QoE), which expands the concept of QoS and describes the user-oriented quality, can directly reflect the degree of recognition of services. As a result, QoE assessment of video services become a research hotspot in the field of network management.In this paper, the research is focusing on the objective QoE evaluation of the video services. Firstly, we analysis the characteristic of the video streaming and abstract the objective factors of the video QoE evaluation, from the bitstream-layer and application-layer. Then we put emphasis on analyzing the application-layer factor-the video content type. At first, for the shorter and little-scene-change videos, this paper presents a fast and effective static video classification algorithm based on cluster analysis. Then, to realize an adaptable video QoE evaluation in real-time environment, we establish a fast scene change detection method and a dynamic content classification mechanism based on the scene change detection.Based on the content types and the objective parameters of bitstream layer and application layer, which have influences on video QoE, a multi-factor QoE evaluation method based on the BP neural network for H.264/AVC encoded video is proposed. Then, combined the dynamic video content classification with video quality evaluation, this paper established a dynamic mechanism for video quality assessment, which realizes the real-time and accuracy of QoE evaluation.Above all, the results of this paper can achieve a certain degree of real time and accuracy on the basis of reducing the cost of time and calculation. We have realized the expected goal in this study.