一种自适应音频双水印算法

An Adaptive Dual Digital Audio Watermarking Algorithm

提出了一种自适应音频双水印算法,算法既能对音频作品的版权进行保护,又能有效地证明其真实性,检测是否遭到篡改.该算法具有以下特点:⑴连续多次嵌入鲁棒水印,以提高水印提取的正确率,增强水印的鲁棒性,嵌入次数根据音频信息量自适应调整;⑵使用不同的量化算法嵌入不同作用的水印;⑶嵌入强度根据音频自身特征选取;⑷在提取水印的过程中不需要原始音频的参与.仿真实验表明:该双水印算法不仅具有较好的不可感知性,而且具有较好的鲁棒性和敏感性.

This paper presents an adaptive dual watermarking algorithm which embeds robust watermark and fragile watermark at the same time.It can protect the copyright of the audio and effectively prove their authenticity.The features of the proposed algorithm are as follows： ⑴ Watermark is embedded several times to enhance the robustness of the watermark.⑵Two different methods are used to embed robust watermark and fragile watermark.⑶The robust watermark is embedded by adaptive quantization according to the audio characteristics.⑷ The algorithm can extract the watermark without the help of the original digital audio signal.Experimental results show that the proposed watermarking scheme is inaudible,robustness and sensitivity.