Study On The Digital Audio Watermarking Algorithm Based On Wavelet Transform Domain

In recent years,due to the development and growth of communication networks and the Internet,the research on copyright protection has gained more new opportunities.Digital watermarking technology provides a more robust solution to copyright protection.Among them,digital audio watermarking technology is an important branch of digital watermarking technology,mainly applied to copyright protection or ownership proof,tamper detection,copy protection,fingerprint,broadcast monitoring and other fields.In this paper,the concept of digital watermarking technology and the characteristics of watermarking in ideal watermarking system are summarized.Then,according to different preconditions,the digital watermarking technology is classified.This paper focuses on the basic principles of digital audio watermarking,several typical attacks,the evaluation methods of audio watermarking and some classical digital audio watermarking algorithms.To improve the performance of the digital audio watermarking algorithm,the following two improved algorithms are proposed:1.Based on discrete wavelet transform,singular value decomposition and independent component analysis,a novel digital audio watermarking algorithm is proposed.Firstly,it preprocesses original audio to select the appropriate fragments to embed watermark,uses three level discrete wavelet transform to the original audio and singular value decomposition to the wavelet transform approximate components.Then it uses a random method to complete mixing to improve watermark's imperceptibility and robustness.When extracting a watermark,it uses discrete wavelet transform and singular value decomposition to the watermarked audio signal.Then it uses Fast-ICA algorithm which based on negative entropy maximization to achieve blind watermark detection.The simulation results show that the algorithm has strong robustness and imperceptibility.2.Combining with discrete wavelet transform and singular value decomposition,a novel digital audio watermarking algorithm is proposed by optimizing of SNR.Firstly,it preprocess original audio to select the appropriate fragments to embed watermark when embedding a watermark,uses three level discrete wavelet transform to the original audio and singular value decomposition to the wavelet transform approximate components,then use optimization of SNR on singular values to improve watermark imperceptibility and robustness.The simulation results show that the proposed algorithm achieve blind watermark detection and has strong robustness and imperceptibility.