.png){kind=icon}
返回
Improved wavelet-based watermarking through pixel-wise masking
.png){kind=icon}
PRE
AI
DOI：10.1109/83.918570.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
A watermarking algorithm operating in the wavelet domain is presented. Performance improvement with respect to existing algorithms is obtained by means of a new approach to mask the watermark according to the characteristics of the human visual system (HVS). In contrast to conventional methods operating in the wavelet domain, masking is accomplished pixel by pixel by taking into account the texture and the luminance content of all the image subbands. The watermark consists of a pseudorandom sequence which is adaptively added to the largest detail bands. As usual, the watermark is detected by computing the correlation between the watermarked coefficients and the watermarking code, anyway the detection threshold is chosen in such a way that the knowledge of the watermark energy used in the embedding phase is not needed, thus permitting to adapt it to the image at hand. Experimental results and comparisons with other techniques operating in the wavelet domain prove the effectiveness of the new algorithm.
Keyword:
image watermarking
perceptual noise masking
wavelets
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
13.7
论文数:
1.0W
被引数:
8.4W
机构
暂无机构信息
引用论文
.png){kind=icon}
Resolving rightful ownerships with invisible watermarking techniques: Limitations, attacks, and implications使用不可见的水印技术解决合法所有权: 限制，攻击和影响
PREAI
分享收藏Orientational dynamics of fluctuating dipolar particles assembled in a mesoscopic colloidal ribbon
OAAI
OAAI分享收藏
查看更多内容