Design Of Phychoacoustic Model Based Audio Watermark In Time Domain

With the emergence of digital era, more and more digital media such as digital TV, digital movie as well as digital music have started to take the place of the traditional analog media. In the mean time, the rapid development of network and portable digital player facilitated the propagation of digital media. However, digital media also have brought about tough challenge for us: the management of the digital copyright. Digital watermark is such kind of technology utilized in the management of the digital media copyright. It embed hidden information of copyright into digital media through information hiding technique which can only be detected with the help of copyright information.In this paper, the development of audio watermark is first introduced. Then, the clarification and the basic process flow are explained. Since the watermark algorithm to be proposed is in the time domain, two prototype time domain audio watermark algorithms are illustrated in detail. These time domain algorithms have 3 defects. That is: they fail to take full adva- antage of human phychoacoustic system to enhance the imperceptibility; they don't include anti synchronous attack methods; their detection result will be distorted by the remain of the correlation. Thus, as to improve tho- se flaws , a new audio watermark algorithm is proposed, it is called: Phyc- hoacoustic Model Based Threshold Predefined Watermark (PMBTPW).In order to take full advantage of human acoustic system, PMBTPW separates the generation of left and right channel watermark in different ways. The SNR of each frequency band is obtained by the analysis of the audio by the combination of wavelet package and phychoacoustic model. Then, watermark is weighted by the SNR so as to make good use of the...