CHEN Liang,LI De*.Robust watermarking based on PN sequence and complete complementary code[J].Journal of Yanbian University,2019,45(03):228-233.
基于PN序列和完全互补码的鲁棒音频水印算法
- Title:
- Robust watermarking based on PN sequence and complete complementary code
- 文章编号:
- 1004-4353(2019)03-0228-06
- 分类号:
- TP391.41
- 文献标志码:
- A
- 摘要:
- 为了解决音频作品容易引发版权纠纷的问题,提出了一种基于PN(Pseudo -Noise)序列和完全互补码的音频扩频水印算法.首先,在时域内引入回声,利用PN序列改变前后向回声内核的极性,并对回声内核进行加密,以此提高水印的安全性和透明性; 然后,利用完全互补码技术对水印序列进行扩频,以此提高水印算法的鲁棒性和嵌入容量; 再次,利用PN序列的短时能量和随机特性,将扩频调制后的水印信息自适应地嵌入到音频信号的时域波形上; 最后,利用PN序列和完全互补码扩频得到水印.实验结果显示,在保证水印透明性的基础上,且在回声嵌入强度小于0.01的情况下,本文提出算法的水印信息检测率超过98%,对高斯噪声、滤波和重采样等攻击的水印信息检测率超过90%,这表明本文算法鲁棒性较强,具有良好的应用价值.
- Abstract:
- In order to solve the copyright disputes of audio works, an audio spread spectrum watermarking algorithm based on PN(Pseudo -Noise)sequence and complete complementary code is proposed. Firstly, echo is introduced in time domain, and the polarity of forward and backward echo cores is changed by PN sequence, and the echo cores are encrypted to improve the security and transparency of watermarking. Secondly, the watermarking sequence is spread spectrum using complete complementary code technology to improve the robustness and embedding capacity of the watermarking algorithm. Thirdly, the short-term energy and random characteristics of PN sequence are used to expand the watermarking algorithm. After frequency modulation, the watermarking information can be adaptively embedded into the time domain waveform of audio signal. Finally, the watermarking is obtained by PN sequence and complete complementary code spread spectrum. The experimental results show that, on the basis of ensuring the transparency of the watermarking, the detection rate of the watermarking information exceeds 98% when the echo hiding strength is less than 0.01, and the detection rate of the assault watermarking information such as adding noise, filtering and resampling exceeds 90%. This shows that the proposed algorithm has strong robustness and good application value.
参考文献/References:
[1] SYHYNDEL P. Watermarking schemes evaluation[J]. IEEE Signal Processing Leters, 2011,17(5):58-64.
[2] 高雪丽,李德.基于边缘特征的动画零水印算法[J].延边大学学报(自然科学版),2017,43(4):353-358.
[3] 杨振仁,李德.基于图像归一化与扩频技术的DTC零水印算法[J].延边大学学报(自然科学版),2017,43(2):179-183.
[4] HU K, KUNDUR D, HATZINAKOS D. Statistical invisibility for collusion resistant digital video watermarking[J]. IEEE Trans on Multimedia, 2015,17(1):43-51.
[5] BHAT H, THENG Y. Robust and inaudible multi-echo audio watermarking[J]. IEEE Pacific Rim Conference on Multimedia, 2002,15(3):713-720.
[6] LEI M, CHENG M, LIU B. An audio zero-watermark scheme based on energy comparing[J]. China Communications, 2011,11(7):110-116.
[7] FALLAHPOUR M, MEGÍAS D. Secure logarithmic audio watermarking scheme based on the human auditory system[J]. Multimedia System, 2014,20(2):155-164.
[8] KHALDI K, BOUDRAA A O. Audio watermarking via EMD[J]. IEEE Transactions on Audio Speech and Language Processing, 2013,21(3):675-680.
[9] 钱清.基于数字水印的语音内容认证算法研究[D].成都:西南交通大学,2018:23-26.
[10] 吕冰.基于线性预测倒谱滤波的音频水印检测技术研究[D].武汉:华中师范大学,2017:19-22.
[11] 马志伟.音频数字水印算法研究与优化[D].哈尔滨:哈尔滨理工大学,2017:22-26.
备注/Memo
收稿日期: 2019-06-10 *通信作者: 李德(1972—),男,博士,教授,研究方向为信息安全与版权保护.
基金项目: 吉林省教育厅“十三五”科学技术研究项目(吉教科合字[2016]第249号)