A fragile digital audio watermarking algorithm based on SMM

A novel digital audio watermarking scheme based on statistical-mean manipulation(SMM) is presented in order to content authentication. The watermarks are embedded at the SMM of the high frequency components of a digital audio signal by making full use of the Multi-resolution DWT. First, the digital audio signal is decomposed by using wavelet. Second, the SMM of the high frequency components of the digital audio signal is computed. Finally, the watermark images after being changed with Arnold shuffling techniques are embedded at the SMM. Experimental results show that the performance of the watermarking scheme is indeed superb in terms of fragility.