摘要
利用音频局部统计特征提出一种可同时抵抗时域拉伸和局部裁剪攻击的音频水印算法.水印嵌入时先对音频信号进行鲁棒自适应分段,然后计算直方图和均值,参照均值修改直方图形状以完成信息隐藏.理论分析和实验结果表明,该方法具有很好的抗时域拉伸性能,且抵抗局部裁剪攻击的性能有显著提高.
This paper proposes a robust audio watermarking approach against random cropping and timescale modifications （TSM） by using two local statistical features.In the embedding,audio signals are divided into segments,and histogram and the absolute mean value of each segment computed.The histogram shapes of the segments are modified to insert messages by referring to the corresponding mean values.Theoretical analysis and experimental results show that the proposed audio watermarking algorithm can provide better performance for random cropping while keeping satisfactory robustness to the TSM attacks in comparison with earlier methods.
出处
《应用科学学报》
CAS
CSCD
北大核心
2014年第4期434-440,共7页
Journal of Applied Sciences
基金
国家自然科学基金(No.60903177
No.61272414)
广州市科技计划重点项目基金(No.2012J410008)资助
关键词
音频水印
鲁棒分段
随机裁剪
局部统计特征
audio watermarking
robust segmentation
random cropping
local statistical feature
