改进型扩频水印算法及在音频认证上的应用

ADVANCED ALGORITHM ON SPREAD-SPECTRUM WATERMARKING AND APPLICATION ON THE AUDIO AUTHORIZATION

提出了一种基于基本音频扩频水印的改进扩频算法,引入随机信号的相关函数μ(x,b)作为水印系数。若使用本算法,加入水印与原始信号的PSNR值为+46.5091dB。本算法与基本扩频算法相比出错概率降低了37个数量级(在SNR=10dB的情况下),能承受的声能量也高出近10dB(在SNR=20dB的情况下),在抗噪性能与对水印能量的要求方面明显有了较大的改善。由于具备扩频水印的优点,因此在对音频的版权认证方面能起到很大的作用。

This paper puts forward the advanced algorithm on spread-spectrum watermarking based on the basic algorithm of spread-spectrum watermarking, which uses the function u（x, b） to be the coefficient of watermarking. If using this algorithm, the PSNR of original signal and signal with watermarking is ＋ 46. 5091. Comparing with the basic spread-spectrum,the error probability has reduced 37order when SNR is 10dB and the bearable disturbance has also increased about 10dB when SNR is 20dB ,which could reduce the disturbance greatly from signal itself and have a large improvement in the bearable load to noise and sensitivity to malicious attacks. Because the algorithm is just improved based on the basic algorithm, it is available for almost system used basic algorithm on spread-spectrum watermarking.