基于提升小波的实时音频数字水印

Audio Digital Watermarking Based on the Lifting Scheme Wavelet Transform

提出一种用于音频数字水印的快速算法。该算法把水印和音频进行分段,水印嵌入在音频小波分解后的低频系数中。为了提高水印的鲁棒性,根据嵌入门限将水印自适应嵌入到音频信号强度大的位置。采用基于提升方案的小波变换,减小水印信息提取的计算量。利用巴克码作为同步信号,用相关法实现实时同步检测。文中给出了详细流程图,对算法复杂度的估计表明,使用现行通用DSP可以实现音频水印的实时提取。实验结果证明了该算法对水印嵌入和提取的有效性和实时提取的可行性。

This paper presents a fast algorithm for audio digital watermarking , which embeds the watermarking into the low frequency coefficients in DWT（Discrete Wavelet Transform ） domain adaptively according to embedding threshold . A fast algorithm for DWT Lifting Wavelet Transform is used for reducing the complexity of watermarking extraction. The Barker code is used for synchronization and the real - time synchronous detection is accomplished through con＇elation. This paper affords detailed flow chart and estimation of the computational complexity which proves that general DSP（ Digital Signal Processors）can accomplish the real - time watermarking extraction. This method is effective in embedding and extracting the watermarking, and also practicable in real time detection according to the results of experiment.