论文部分内容阅读
图像隐形水印是解决数字媒体版权保护的有效手段 .目前的大多数水印算法采用伪随机序列构成水印 ,因而只能给出“Yes”(有水印 )或“No”(无水印 )的回答 .嵌入有意义的水印在一些应用场合更加重要 ,但同时也带来了更多的困难 .嵌入有意义的水印必须解决二个问题 :(1)嵌入的水印具有一定的数据量 ;(2 )从有失真的水印图像中正确地提取嵌入的水印 .本文应用数字通信的理论和方法 ,提出了一个嵌入有意义字符串的DCT域隐形水印算法 .该算法采用BCH编码和匹配滤波来降低误码率 .应用所提出的算法 ,把一个 32个字符组成的水印嵌入到 2 5 6× 2 5 6的图像中并达到了较好的稳健性 .在水印图像信噪比受Gaussian加性噪声干扰下降到 2 6dB时 ,仍实现了对嵌入水印零错误概率的检测 .
Image invisible watermarking is an effective method to solve copyright protection of digital media.Most of current watermarking algorithms use pseudo-random sequence to form watermark, so only “Yes” or “No” Meaningful watermarking is more important in some applications, but it also brings more difficulties at the same time.Embedding a meaningful watermark must solve two problems: (1) the embedded watermark has a certain amount of data; (2) Distortion watermarking image.According to the theory and method of digital communication, this paper proposes a DCT invisible watermarking algorithm which embeds meaningful strings.This algorithm uses BCH coding and matched filtering to reduce the bit error rate. The proposed algorithm embeds a 32-character watermark into 256 × 256 images and achieves good robustness.When the watermark image SNR is reduced to 2 by Gaussian additive noise 6dB, still realize the zero probability of embedded watermark detection.