【24h】

Automatic objective picture quality measurement method using invisible marker signals as Reduced References

机译:使用不可见标记信号作为简化参考的自动客观图像质量测量方法

获取原文
获取原文并翻译 | 示例

摘要

To automate the operation and monitoring of digital video transmissions, the authors propose an objective picture quality measurement method for coded pictures that uses embedded invisible marker signals and does not require reference pictures. The authors previously proposed a method for estimating PSNR by using the rate of false detection of binary information expressed by the marker signals. This paper proposes another method for estimating PSNR. This method is based on direct measurement of the magnitude of the marker signal degradation instead of the false detection rate of the embedded information. Since the theoretical model shows that the PSNR of the decoded pictures can be approximately calculated by measuring the degradation of only one amplitude component of the spread spectrum, the marker is embedded into the amplitude component of the spread spectrum domain and is detected from the decoded pictures to calculate the degradation. The proposed method can be applied to various types of video transmissions because this method is valid regardless of the kind of pictures, bitrates, or number of tandem codec connection stages. Furthermore, the proposed method is also applicable to high quality video transmissions because the degradation caused by embedding markers is negligibly small.
机译:为了使数字视频传输的操作和监视自动化,作者提出了一种针对编码图像的客观图像质量测量方法,该方法使用嵌入的不可见标记信号,并且不需要参考图像。作者先前提出了一种通过使用标记信号表示的二进制信息的错误检测率来估计PSNR的方法。本文提出了另一种估计PSNR的方法。该方法基于标记信号衰减幅度的直接测量,而不是嵌入信息的错误检测率。由于理论模型表明,仅通过测量扩展频谱的一个幅度分量的降级就可以近似计算出解码图像的PSNR,因此将标记嵌入扩展频谱域的幅度分量中,并从解码图像中进行检测计算退化。所提出的方法可以应用于各种类型的视频传输,因为该方法是有效的,而与图片的类型,比特率或串联编解码器连接阶段的数量无关。此外,所提出的方法还适用于高质量视频传输,因为由嵌入标记引起的降级可以忽略不计。

著录项

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