New adaptively quantized audio watermark algorithm based on wavelet transform

A digital audio watermarking algorithm based on adaptive quantization is presented. The algorithm modifies the quantization step by adaptively adjusting the quantified SNR and determines the intensity of the embedded watermark according to features of the audio signal. As a result, watermark extraction does not require knowledge of the original audio signal, allowing blind extraction of the watermark. Simulation results show that the proposed watermarking scheme is not only transparent, but also robust against general signal-processing operations such as noise addition, lossy compression, low pass filtering, and resampling.