首页> 外文会议>International Conference on Virtual Systems and Multimedia(VSMM2004); 20041117-19; Ogaki City(JP) >Watermarking Method to Embed Bitstream onto Singular Values Given by Singular Value Decomposition
【24h】

Watermarking Method to Embed Bitstream onto Singular Values Given by Singular Value Decomposition

机译:通过奇异值分解给出的比特流嵌入奇异值的水印方法

获取原文
获取原文并翻译 | 示例

摘要

Singular value decomposition (SVD) decomposes a matrix into two orthogonal matrixes and a diagonal matrix. This diagonal matrix has a property such that it is invariant to the addition of linear independent row/column to original matrix, which is a very convenient property for the watermarking. We propose the watermarking method which embeds the bitstream by giving a slight modification to the singular values and confirm its robustness against various image processing. The proposed method has the following properties. (1)It does not need the original image to extract the watermark. (2)It is robust against random number addition and low bit deletion because the watermark is diffused to allover image. (3)It gives a convenient way for calculating SNR.
机译:奇异值分解(SVD)将矩阵分解为两个正交矩阵和一个对角矩阵。该对角矩阵具有使得向原始矩阵添加线性独立行/列不变的特性,这对于加水印而言是非常方便的特性。我们提出了一种水印方法,该方法通过对奇异值进行一些修改来嵌入比特流,并确认其对各种图像处理的鲁棒性。所提出的方法具有以下特性。 (1)提取水印不需要原始图像。 (2)由于水印被扩散到整个图像上,因此对于随机数增加和低位删除是鲁棒的。 (3)为计算信噪比提供了一种方便的方法。

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