一种新颖的双声道数字音频无源水印算法

A New Algorithm with Two Channel Blind Digital Audio Watermark

提出了一种新颖的双声道数字音频无源水印算法。本算法先对需要嵌入的灰度水印图片做置乱加密处理,同时对音频左右声道分别进行三级小波分析,在三级小波低频系数中自适应地寻找左右声道的差异点作为嵌入位置。并能自适应地以水印图像大小为选择基准,将低频系数差异点进行分段。最后,根据基准值和差异参数将水印嵌入到低频系数中。实验结果表明该算法简单易行,嵌入水印后的音频几乎不能被感知改变,安全性和隐蔽性高,并且提取水印图像时无需原始音频。本方法新颖之处在于:能大幅提高嵌入水印信息量,对于灰度以及彩色图像同样适用;具有很好的隐藏性,嵌入水印后的音频难以感知水印的存在;以水印图像像素为选择基准能自适应地将水印嵌入到音频中。

As a popular copyright protection technology,the digital watermark technology has been cause for concern in recent years. This thesis introduces a new algorithm with two channels blind digital audio watermark.At first,the system encrypts and scrambles watermark-image.At the same time,the host audio is decomposed by 3-level wavelet transform.The system is able to choose the deferent point between L-channel and R-channel as the best insert position from the low frequency coefficients by itself.These low frequency coefficients segmente based on the size of watermark-image adaptively.Finally,the watermarking is inserted in the audio by them.The result shows,it is hard to find the different between the original audio and watermarked audio.This system is easy to operate and the watermark can be extracted without original data.The Novelty of this algorithm is：the system supports insert a high-capacity Gray scale and color watermark images;It does well in transparency.It is hardly to feel if the audio has been inserted a watermark; System can insert the watermark-image by itself base on the attribute of image＇s element.