基于DWT-SVD的大容量音频水印算法

在音频水印领域,嵌入大量水印而不影响宿主音 频的听觉质量,同时具有良好的鲁棒 性,仍然是一项具有挑战的工作。针对这一问题,本文提出了一种大容量,透明,鲁棒的音 频水印算法。首先对音频应用2级离散小波变换(DWT),并选取低频分量作奇异值分解(SV D),然后将水印嵌入在奇异值矩阵的非对角线元素中,可同时提高算法的透明性和嵌入容 量。另外,利用汉明码编码提升水印的鲁棒性和安全性。仿真实验结果表明,所提算法在平 均信噪比37.75 dB的情况下,可以达到5512 bps 的嵌入容量,对噪声添加、滤波、剪切、压缩等各种信号处理攻击具有良好的鲁棒性。

Large capacity audio watermarking algorithm based on DWT-SVD

In the field of audio watermarking,reliably embedding the large numbe r of watermarking bits per second in an audio signal without affecting the audib le quality of the host audio with good robustness against signal processing atta cks is still a challenging work.To solve this problem,a high payload,transpar ent and robust audio watermarking algorithm based on discrete wavelet transform (DWT) and singular value decomposition (SVD) is proposed in this paper.First of all,the host audio signal is transformed to DWT and the low-frequency compone n ts are selected.The low-frequency components of the host audio have been selec t ively embedded with watermark image making the watermarked signal highly imperce ptibility and robustness.Subsequently,the low-frequency components is perform e d SVD and watermark image being embedded into the non-diagonal elements of the s ingular value matrix after SVD,which can simultaneously improve the transparenc y and capacity of the proposed algorithm.In addition,the hamming code encoding is used to improve the robustness and security of the watermark.Experimental r esults show that the proposed audio watermarking algorithm can achieve 5512bps capacity at an average signal-to-noise ratio (SNR) of 37.75dB with good robustn ess to various signal processing attacks such as noise addition,filtering,crop ping and compression.The comparison of the proposed algorithm with the other DW T-SVD techniques also shows that the proposed algorithm is able to achieve hig h payload with good robustness under perceptual constraints.