Digital Audio Watermarking Algorithm Based On Hybrid Domain

In recent years,the Internet and multimedia technologies have a escalating development,and various forms of multimedia digital products can be easily distributed,disseminated and acquired on the Internet.Among them,digital audio products are the most commonest to multimedia products.However,this has provided opportunities for unauthorized transmission and acquisition.Illegal copying,illegal falsification and illegal dissemination of multimedia products have caused piracy.Therefore,the research on information security of multimedia digital products,especially audio digital products,is important and urgent.After years of development and research,audio digital watermarking technology has gradually become the most effective means of copyright protection for the field of multimedia information security.This technique embeds information that needs to be kept secret in the host carrier signal without affecting the signal quality of the host carrier,and hides its existence.For the copyright protection for audio multimedia information,two audio digital watermarking algorithms based on hybrid transform domain are proposed to this paper.One is to propose a hybrid domain audio digital watermarking algorithm that combines redundant discrete wavelet transform,discrete cosine transform and QR decomposition.In the algorithm,redundant discrete wavelet transform is performed on the segmentation of the original audio carrier signal,discrete cosine transform is performed on the transformed components,and the converted low frequency coefficients are converted into square matrixes and QR decomposition is performed.The ratio between the first two values S(1,1)and S(2,2)of the triangular matrix embeds the watermark information in S(1,1).The watermark is embedded in the approximation component of the redundant discrete wavelet transform.This algorithm not only has good robustness,but also guarantees the embedding capacity of the watermarking.The QR decomposition has good stability,so the algorithm can resist a variety of common audio signal processing attacks.The second is to propose an audio digital watermarking algorithm with distortion compensation based on discrete wavelet transform and discrete cosine transform.In the algorithm,the original audio carrier signal is segmented,each component of the audio is subjected to discrete cosine transform on the discrete wavelet transform,and the low frequency part of the obtained transform coefficient is quantized according to the distortion-compensated jitter quantization modulation method to complete the watermark information.Embedded.The distortion compensation for the algorithm can reduce the signal distortion caused by quantization.Experimental results show that these two hybrid domain algorithms have strong resistance to common attacks and signal processing operations such as MP3 compression,resampling,low-pass filtering,re-quantization,and Gaussian white noise,compared to single transform domain,they have more advantage on the field of robustness and imperceptibility.