• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文学位 >Real-time video quality control for multimedia network.
【24h】

Real-time video quality control for multimedia network.

机译：多媒体网络的实时视频质量控制。

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In this thesis, we propose a new approach for video quality control for multimedia networks. Our new approach is based on video quality measure that combines both the network quality of service (QoS) as well as the user quality of experience (QoE). The proposed approach improves the end-to-end traditional video quality control for multimedia network by including the human perception of video data, which is major concern for the video client, along with the network quality of service (QoS) measurements. In our approach we use packet loss rate as quality of service (QoS) parameter, and self-reference complex wavelet video structural similarity index (SRCW-VSSIM) as quality of experience (QoE) parameter. Compared with traditional QoS only video quality control technique, the proposed video quality control technique for multimedia networks is based on including both QoS and QoE parameters. We will show that the proposed QoS-QoE based video quality control algorithm can reflect both the condition of the network environment and the human perception of the received networked video data stream. According to both QoS and QoE parameters, rather than using only QoS parameter, video quality control action will satisfy the user needs more than relying only on the network conditions. Since our proposed QoE parameter SRCW-VSSIM can be obtained with no reference (NF) video data, it satisfies the requirement of real-time video transmission.;In addition to our video quality control technique, we introduce machine learning approaches for combined QoS-QoE based video quality control techniques for real-time streaming service. The proposed schemes are based on statistical learning technique, Support Vector Regression (SVR), to predict combined QoS-QoE parameter in the near future. The character of machine learning technique makes this scheme proactive, and be able to trigger the rate control action to adjust the video streaming rate before network conditions start deteriorating. QoS-QoE based video quality control indicator (QQVQCI), defined as the combined QoS-QoE parameter for real-time video quality control, mixed with QoE index are used to generate training dataset to predict QQVQCI in the near future.;Theoretical analysis as well as simulation results are presented.
机译：本文提出了一种新的多媒体网络视频质量控制方法。我们的新方法基于视频质量度量，该度量结合了网络服务质量（QoS）和用户体验质量（QoE）。所提出的方法通过包括人类对于视频客户端的关注的视频数据感知以及网络服务质量（QoS）测量，改善了多媒体网络的端到端传统视频质量控制。在我们的方法中，我们使用丢包率作为服务质量（QoS）参数，并使用自参考复数小波视频结构相似性指标（SRCW-VSSIM）作为体验质量（QoE）参数。与传统的仅QoS视频质量控制技术相比，所提出的多媒体网络视频质量控制技术基于QoS和QoE参数。我们将显示，提出的基于QoS-QoE的视频质量控制算法既可以反映网络环境的状况，又可以反映人对接收到的联网视频数据流的感知。根据QoS和QoE参数，而不是仅使用QoS参数，视频质量控制操作将满足用户需求，而不仅仅是依靠网络条件。由于我们提出的QoE参数SRCW-VSSIM可以在没有参考（NF）视频数据的情况下获得，因此它满足实时视频传输的要求。；除了我们的视频质量控制技术外，我们还介绍了结合QoS的机器学习方法基于QoE的实时流服务的视频质量控制技术。所提出的方案基于统计学习技术，支持向量回归（SVR），以在不久的将来预测组合的QoS-QoE参数。机器学习技术的特性使该方案具有主动性，并能够在网络状况开始恶化之前触发速率控制动作来调整视频流速率。基于QoS-QoE的视频质量控制指标（QQVQCI）被定义为用于实时视频质量控制的QoS-QoE组合参数，并与QoE指标混合，用于生成训练数据集以预测QQVQCI在不久的将来。以及仿真结果。

著录项

  • 作者

    Jiang, Biao.;

  • 作者单位

    The City College of New York.;

  • 授予单位 The City College of New York.;
  • 学科 Computer engineering.;Information Technology.;Multimedia.
  • 学位 Ph.D.
  • 年度 2013
  • 页码 89 p.
  • 总页数 89
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号