Research And Optimization Of Audio Digital Watermarking Algorithm

In recent years,with the development of Internet and multimedia,digital audio products become more and more popular,audio communication becomes simple,which also provides an opportunity for unauthorized transmission.Audio digital copyright protection to become a problem to be solved.Audio digital watermarking technology is an important way to solve the audio digital copyright protection.In order not to affect the normal use of audio,you need to make the audio digital watermark has a good imperceptibility.Audio in the transmission process,there may be noise,compression,filtering and other effects,these effects may be to a certain extent,damage the watermark,so a good sound digital audio watermark has important research value.In this paper,the imperceptibility and robustness of audio digital watermarking are restricted,and the selection of watermarking algorithm parameters,the spread-transform dither modulation algorithm and the singular value decomposition algorithm are studied,and apply it to the embedding and extracting of audio digital watermarking.The optimization algorithm is applied to digital watermarking more and more widely.Particle swarm optimization and genetic algorithm are studied and improved,which can be used to optimize the audio digital watermarking algorithm and improve the performance of the algorithm.The particle swarm optimization algorithm is improved to make it converge faster and not to fall into local optimum.The main work of this paper is as follows:1.An audio digital watermarking algorithm and optimization scheme based on spread-transform dither modulation and particle swarm optimization are proposed.The spread-transform dither modulation algorithm is applied to the audio digital watermarking,and the embedded digital watermarking and blind extraction are realized by the spread-transform dither modulation algorithm.Particle swarm optimization algorithm is applied to the spread-transform dither modulation algorithm.The three levels of discrete wavelet transform is performed on the carrier audio,and the discrete cosine transform is performed on the low frequency components.The watermark embedding is performed in the mixed domain,and the embedding position is selected adaptively according to the embedded information amount.The quantization step size and the number of projection elements of the spread-transform dither modulation algorithm are optimized by particle swarm optimization algorithm,and the contradiction between imperceptibility and robustness is weighed.2.An audio digital watermarking algorithm and optimization scheme using singular value decomposition and genetic algorithm are proposed.The singular value decomposition algorithm is applied to audio digital watermarking to realize the embedding and blind extraction of audio digital watermark.Read the carrier audio file,according to the size of the embedded image,select the large value of the data,the large-dimensional data into two-dimensional matrix.The three-dimensional discrete wavelet transform is applied to the two-dimensional matrix,select the low-frequency components for singular value decomposition,select the large value in the singular value for watermark embedding.Genetic algorithm is improved in genetic algorithm to improve the convergence of genetic algorithm and expand the search space.The improved genetic algorithm is applied to optimize the embedding strength of audio digital watermarking,and the contradiction between imperceptibility and robustness is balanced.Experiments show that on the basis of weighing robustness and imperceptibility,the algorithm of the third chapter has a stronger robustness,the algorithm of the fourth chapter has a stronger imperceptibility,can choose the appropriate algorithm according to your specific needs.This article has the practical significance to perfect the audio digital watermarking theory and solve the digital copyright protection of the audio.