Adaptive audio watermarking algorithm based on SNR in wavelet domain

An algorithm of additive audio watermarking based on SNR to determine a scaling parameter a is proposed in this paper. The visually recognizable binary watermark image could be embedded in the original signal wavelet domain in the algorithm. The intensity of embedded watermarks on the different audio segments can be modified by adaptively adjusting the scaling parameter alpha. Comparing to the algorithm without using SNR, the experimental results demonstrate that the watermark is inaudible and the algorithm is robust to many operations of digital audio signal, such as resampling, cropping, low pass filtering, additive noise and so on.