Advanced Wigner Method Used in Progress of Embedding Watermark Message into Audio File

Abstract

In this paper, we are supposed to use an advanced Wigner method during the procedure of embedding watermark into audio file. The watermark message is embedded in the most significant place in original audio file. By decomposing the audio file into smaller frequency band using wavelet transform, the watermark message can be included in the coefficients. The watermarked audio signal is obtained by wavelet reconstruction. The advanced Wigner method, which is based on STFT (short time Fourier transform) and Wigner, takes advantage over other methods. It considers the most significant frequency band and certifies the watermark existing in the product with ease. Using wavelet transform makes good use of the multi-resolution characteristics of it. During detecting the watermark message, the original audio signal is not needed, just only watermark keys are used. The simulation results show that the algorithm is inaudible and robust to strong attacks such as noise addition, filtering and re-sampling. However, with the MP3 compression, the algorithm does not work well.