Multifunctional Audio Digital Watermarking Based On Psychoacoustic Model Study

In modern information society, the digital audio and internet technology is developing rapidly. When we are enjoying the extensive and convenient digital life bought by modern information technology, the security problem of digital media especially of digital audio has been coming out. Some vicious individuals and groups may copy, tamper, diffuse the copyrighted audio digital works without the permission of the owners, which makes huge damage to the owners and publishers of the works.In this case, audio digital watermark tehnology as an effect method to solve the copyright problem of the digital audio works, will play a important role on protecting the works' copyright. This article presented a multifunctional audio digital watermark algorithm based on psychoacoustic model. With masking effect of sound, This algorithm calculated a masking threshold through psychoacoustic model, and made audio sound's intensity lower than the masking threshold to make watermark more imperceptible. Before embedding the watermark, the original audio signal was divided into several frames and FFT was performed first, then we embedded two watermarks into the carrier audio signal (multifunctional audio digital watermark): a robust watermark in the low frequency region and a fragile watermark in the high frequency region, this is the biggest innovation of this article. The robust watermark could resist vicious attack such as signal processing, and could be used to affirm the owner of the copyright, the fragile watermark had frangibility to some signal processing, so we could use it to authenticate it's integrality. Finally, we extracted the watermark through the counter-process of the embedding.After embedding the watermark, this article discussed the resistance against the attacks such as MP3 lossy compression, noising adding,low pass filtering and resampling. As the experimental results showed, the robust watermark based on the algorithm was apparently similar to the original watermark image, and was so highly robust that it could affirm the copyright effectively, the fragile watermark had such good sensitivity to those attacks that it could make sure that if the signal had been tampered or destroyed., so we can authenticate the integrality of the audio works.