Audio Information Hiding

Information hiding is a new research domain. It hides a secret object called embeddedobject into another exoteric object called cover object and gets the stego object whichcontains the secret object. The extractor does not know whether the cover object containssecret object or not, even if the extractor knows, there is hard to extract or demolish thesecret information. The cover object could be document, picture, audio, video and soon.To compare with the encryption technology, it adds another protection, that is becomethe information which needs to be protected from "do not understand" to "can not see".The thesis is on digital audio information hiding, it contains five parts:Firstly, it reviews the evolution of information hiding, describes some basic items ofinformation hiding and digital audio information hiding.Secondly, combine the Human Audio System and masking effect, it introduces thebasic knowledge of digital audio processing and the traditional methods on audioinformation hiding, and carries on the analysis and comparison to these methods. Then itlists some requisitions and function evaluation methods on digital audio informationhiding.Thirdly, according to the analysis on digital audio information hiding algorithms inthe wavelet domain, with the theory of image scrambling and chaotic sequence, a newdigital audio information hiding algorithm in DWT is proposed, well made full use of theproperties of chaotic sequence which are the initial value sensitivity, Gaussian Noiseresembling and low correlation.Fourthly, in consideration of various attacks that the audio cover object would besubjected to, the above algorithm is improved to meet the requirements of security androbustness. The quantization and imitate quantization algorithms are both better than theabove algorithm, but, the imitate quantization algorithm is much better than quantizationalgorithm in invisibility and imperceptibility.Finally, we give simulated results about the algorithms, and made an analysiscomparison.