An Audio Mask Algorithm Based on Cepstrum Tansform and Wavelet Tansform

Abstract

In order to better improve audio watermark robustness and perceptual, the paper puts forward a kind of based on cepstrum transform and the wavelet transform of digital audio watermarking algorithm The algorithm will be audio signal wavelet decomposition, selection of low frequency coefficient of framing and cepstrum transform And according to the statistical mean thought, using the cepstrum domain characteristics of the energy concentration near zero, a pseudo random sequence watermark embedded in cepstrum domain experimental results show that the algorithm not only has good perceptual, but also has better robustness and can withstand attacks of common signal such as low-pass filter resampling quantitative lossy compression and sync to attack.