Wu Shiqiang, Huang Ying, Guan Hu, Zhang Shuwu, Liu Jie
School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing 101408, China.
Institute of Automation, Chinese Academy of Sciences, Beijing 100045, China.
Entropy (Basel). 2022 Dec 17;24(12):1843. doi: 10.3390/e24121843.
Digital audio watermarking is a promising technology for copyright protection, yet its low embedding capacity remains a challenge for widespread applications. In this paper, the spread-spectrum watermarking algorithm is viewed as a communication channel, and the embedding capacity is analyzed and modeled with information theory. Following this embedding capacity model, we propose the extended-codebook spread-spectrum (ECSS) watermarking algorithm to heighten the embedding capacity. In addition, the diversity reception (DR) mechanism is adopted to optimize the proposed algorithm to obtain both high embedding capacity and strong robustness while the imperceptibility is guaranteed. We experimentally verify the effectiveness of the ECSS algorithm and the DR mechanism, evaluate the performance of the proposed algorithm against common signal processing attacks, and compare the performance with existing high-capacity algorithms. The experiments demonstrate that the proposed algorithm achieves a high embedding capacity with applicable imperceptibility and robustness.
数字音频水印是一种很有前途的版权保护技术，但其较低的嵌入容量仍然是广泛应用面临的一个挑战。本文将扩频水印算法视为一种通信信道，并运用信息论对其嵌入容量进行分析和建模。基于该嵌入容量模型，我们提出了扩展码本扩频（ECSS）水印算法以提高嵌入容量。此外，采用分集接收（DR）机制对所提算法进行优化，在保证不可感知性的同时，获得高嵌入容量和强鲁棒性。我们通过实验验证了ECSS算法和DR机制的有效性，评估了所提算法抵御常见信号处理攻击的性能，并与现有的高容量算法进行了性能比较。实验表明，所提算法在具有适用的不可感知性和鲁棒性的情况下实现了高嵌入容量。
