摘要
用户对视频的直观体验是流畅度和清晰度,基于vMOS的评测标准,将网络侧和用户侧的因素进行综合优化将是难中之难,在不断的发展过程中,会产生更多的影响因素,加之与反馈模块进行结合,更加增加了优化难度。本文首先通过单片机与网络和用户端之间的数据传送来实现反馈的功能。其次将用户观看视频的感受进行数字化处理,并对影响视频指标vMOS的各种因素进行分析,特将vMOS的得分细化并且通过单片机和服务器之间的自动优化功能进行调节处理。最后对各因素与vMOS值之间的相关性进行分析。在排除相关性较低的变量之后,分别得到了sLoading和sStalling这两个变量,通过Matlab进行数据的初步拟合之后得到了一定的函数关系。从而实现了用单片机和服务器对影响因素进行优化,同时优化了视频质量。
The user experience of video is fluency and clarity. The network side and user side factors of comprehensive optimization will be difficult based on the assessment of vMOS. The process of unceasing development will produce more influencing factors, and combined with feedback module, the difficulty of optimization has increased. Firstly, this paper will realize the function of the feedback through the single chip microcomputer and data transfer between client and network. Secondly, it will digitize the feeling of the user to watch video and analyze the various factors influencing the video index vMOS. The vMOS score is refined and the automatic optimization function between the MCU and the server is adjusted. Finally, the correlations between various factors and vMOS value were analyzed. After eliminating low correlation variable, respectively two variables sLoading and sStalling, the initial fitting function is obtained by Matlab for data. So it has realized the optimization of the influence factors, using single chip microcomputer and the server at the same time to optimize the video quality.
出处
《软件工程与应用》
2018年第1期53-59,共7页
Software Engineering and Applications
基金
国家自然科学基金项目(青年科学基金项目)71602124
辽宁省自然科学基金项目20170540662
辽宁省教育厅项目(高等学校基本科研项目青年项目)LQN201710
辽宁省大学生创新项目(沈阳工程学院)。
