Audio Watermarking Algorithm Based On Chaos Encryption In DWT Domain

With the rapid development of computer network and information technology,the expansion of the Internet is greatly accelerated, on the one hand it has madeoutstanding contributions for the entire human social progress and development; onthe other hand accompanied by the development of the Internet and a wide range ofapplications, there are some relevant issues, such as security problems ofmultimedia content, which seriously affect the process of development of digital, ifthis problem can not be resolved, the entire electronic publishing industry (includingbooks, music, movies, etc.) will difficult to obtain healthy development. Audiodigital watermarking technology enables audio copyright protection, throughembedding the meaningful watermark information into the audio file to achieve theprotection of copyright and confidential communications purposes. This articledescribes the digital audio watermarking technology, use of wavelet transform,wavelet packet transform, discrete cosine transform technology to complete theaudio watermark embedding and blind extraction, the text of the main tasks are asfollows:(1) Detailed exposition of the concept of audio digital watermarking,performance requirements, describes the characteristics of the human auditorysystem especially masking model, summarized the classic audio watermarkingalgorithm, given common evaluation criteria for audio watermarking.(2) Use Arnold scrambling and chaotic encryption as dual Encryption means ofwatermark information pretreatment, Arnold scrambling is used to change theposition of the watermark image pixel, then the pixel values are processed usingchaotic encryption though the illegal has extracted watermark, it is difficult to get anaccurate watermark information, greatly improving the confidentiality and securityof the watermark.(3) This article proposes two blind audio watermarking algorithms. One is anaudio watermarking algorithm based on the comparison of energy in DWT domain,this algorithm combines wavelet transform with discrete cosine transform, selectsembedded position based on the energy, completes the blind watermark embeddingand extraction process; the other algorithm is designed for vinyl records and otherhigh-quality music in wav format. The algorithm combines the perceptualcharacteristics of the human ear to establish a Bark scale wavelet packetdecomposition model, after considering select the appropriate transform coefficientsto quantify to embed the watermarking. There are two scenarios，the former requiresthe watermarking is robust, while the latter has more requirements for watermarking imperceptibility.