摘要: 提出一种基于离散余弦变换(DCT)和奇异值分解(SVD)的音频水印算法。该算法将原始音频进行分段后再分块,对每一块进行DCT变换,利用Zig-Zag算法将变换后的系数分成4个象限,通过对每个象限进行SVD变换得到对角阵,并在对角阵的第1个元素内嵌入水印。实验结果证明,该算法具有良好的透明性,在嵌入强度为0.2时峰值信噪比较高,同时可以抵抗重采样、回声、噪声等攻击,具有较高的鲁棒性。
关键词:
音频水印,
离散余弦变换,
奇异值分解,
透明性,
鲁棒性
Abstract: This paper proposes an audio watermarking algorithm based on Discrete Cosine Transform(DCT) and Singular Value Decomposition(SVD). An original audio is split into fragments, and each fragment is divided into blocks. Each block is decomposed on DCT transformation, whose coefficients are divided into four quadrants in accordance with Zig-Zag algorithm. After each quadrant is decomposed on SVD, a diagonal matrix is obtained. The watermarking information is embedded into the the first element of diagonal matrix. Experimental results show that the algorithm has good transparency, higher PSNR when the embedding strength is 0.2, and it has strong robustness against the audio signal processing such as the reshampling, echoes, noise attack and so on.
Key words:
audio watermarking,
Discrete Cosine Transform(DCT),
Singular Value Decomposition(SVD),
transparency,
robustness
中图分类号:
朱宪花, 雷敏, 杨榆, 王国鑫. 一种基于DCT和SVD的音频水印算法[J]. 计算机工程, 2012, 38(19): 111-113,118.
SHU Xian-Hua, LEI Min, YANG Yu, WANG Guo-Xin. An Audio Watermarking Algorithm Based on DCT and SVD[J]. Computer Engineering, 2012, 38(19): 111-113,118.