基于离散小波变换的盲水印研究与仿真

Research and Simulation on Blind Watermarking Based on DWT

研究优化水印安全性问题。由于水印图像应具备版权信息和密钥条件,针对目前音频水印算法中水印容易被提取,且检测水印时均需要原始水印图像,导致水印安全性,鲁棒性无法保证。提出了基于离散小波变换的混沌加密的音频数字水印盲算法。将混沌原理引入音频水印中,改善了水印的安全性,同时根据信号的能量对水印进行了量化嵌入,改善了水印的鲁棒性。首先利用密钥对二值水印图像进行随机置乱;然后将经过随机置乱处理的含有版权信息的二值水印图像嵌人到原始音频信号中。在水印检测过程中利用已知的密钥对水印进行重构。仿真证明,在不知道密钥的情况下,无法得出水印图像,为设计提供了依据。

This paper researches optimization problem of watermarking security.Because of the shortcomings that in current audio algorithm,the watermark is easy be extracted and watermark detecting requires original watermark images,which can not guaranteed to the watermark security and robustness.The paper presents an audio digital watermark algorithm of discrete wavelet transform based on chaotic encryption.Chaos theory of audio watermarking algorithm was introducing to improve the security of watermark,and the energy of the signal was quantified on the embedded watermark to improve the robustness of the watermark.First,a key was used to randomly scramble the binary watermark image;and then it is embedded to the original audio signal.In the watermark detection process,the known key was used to reconstruct the watermark.Thus,the watermark image can not be obtained without the key.