鲁棒的DCT域音频盲水印算法

Robust blind audio watermarking algorithm in DCT domain

选择合适的DCT系数嵌入水印,可以使鲁棒性和不可感知性达到很好的平衡。为了找到这个系数,利用人耳听觉的频率掩蔽效应分析了DCT系数的听觉容差,发现直流系数具有最大的听觉容差,非常适合嵌入水印。提出了一种新的通过改变DCT直流系数的音频盲水印算法。实验结果表明,嵌入水印后的音频文件具有良好的不可感知性,与其它DCT算法得到的水印文件比较具有更强的鲁棒性,能够抵抗噪声、低通滤波、重采样、回声和Mp3压缩等的攻击。

Selecting an appropriate DCT coefficient to embed in watermark can make the robustness and the imperceptibility reach a good balance, In order to find this coefficient, the auditory masking property was used to analyze the influences of the changing DCT coefficients to the hearing, A good property is found that the human auditory system has the biggest hearing margin to the changing value of the DC coefficient. So the DC coefficient is quite fit for watermarking. Based on the conclusion, a new robust blind audio digital watermarking algorithm by changing the value of the DC coefficient is proposed. Experimental results show that the watermarked audio has good imperceptibility. By contrast with other algorithms in DCT domain, this algorithm has better robust performance, which is robust against different kinds of attacks, such as noise, low-pass filtering, resampling, echo and Mp3 compression.