摘要
为了提高数字音频水印中的水印嵌入效率,利用同时计算N点实序列的离散傅立叶变换(DFT)和逆离散傅立叶变换(IDFT)的新公式提出了一种改进的数字音频水印嵌入方案。通过对DFT的一些性质的分析,利用DFT的对称性,给出了同时实序列的DFT和IDFT的新公式,并改正了Gunther相关公式中的少数错误。在此基础上,将新公式应用到数字音频水印的嵌入算法中,通过实验表明了新方案大约可提高30%的水印嵌入性能。
In order to improve the efficient of digital audio watermarking of simultaneously calculating the discrete Fourier transform and inverse embedding , by exploiting the new formulas discrete Fourier transform of a N-real sequence, an improved digital audio watermarking embedding scheme was proposed. From the analysis of DFT's properties and the syrmnetry of DFT, the new formulas of simultaneously calculating the DFT and IDFT of a real sequence were obtained. The new formulas were applied to digital audio watermarking embedding algorithm. Experiments showed that the new scheme can increase the watermarking embedding oerformance bv 30%.
出处
《四川大学学报(工程科学版)》
EI
CAS
CSCD
北大核心
2008年第2期96-100,共5页
Journal of Sichuan University (Engineering Science Edition)
基金
现代通信国家重点实验室基金资助项目(51436010
505SC010)
四川省教育厅自然科学项目青年基金资助项目(2006B051)
关键词
N-点实序列
离散傅立叶变换
逆离散傅立叶变换
数字音频水印
real N-point sequence
discrete Fourier transform
inverse discrete Fourier transform
digital audio watermarking