Research On Audio Information Hiding

With the rapid development of the Internet,multimedia files can be simply and quickly transmitted,which has brought great convenience to human beings,and also brought great challenges to information security and copyright protection.The traditional encryption technology transforms the information into unrecognizable cipher text,but this can easily attract the attention of attackers and be more easily cracked.As an important branch of confidentiality,audio information hiding technology hides encrypted information in audio files,which not only hides the content of encrypted information,but also the existence of encrypted information.Because of the advantages of audio information hiding technology,it has attracted more and more scholars' attention,and has been widely used in copyright protection,covert communication,etc.Therefore,audio information hiding technology has important research value and broad application prospects.This paper first reviews the audio information hiding technology.Then,based on the existing problems of audio information coding algorithms,combining MDCT,non-uniform quantization and other ideas,this paper proposes an audio information coding algorithm based on MDCT.In order to further improve the robustness of the algorithm,a differential decoding algorithm based on spread spectrum is proposed.The main work of this article includes:1.First,the relevant knowledge of audio information hiding technology and psychoacoustic model is summarized,the system framework,performance indicators and classic algorithms of audio information hiding technology are introduced,then the psychoacoustic model and human auditory system are introduced;finally,the paper introduce the principle of MP3 encoding.By researching and analyzing the existing basic knowledge,it lays the foundation for the proposed audio information encoding and decoding algorithms.2.The paper proposed an audio information coding algorithm based on MDCT.In order to make up for some problems of FFT and DCT,this algorithm chooses MDCT as the time-frequency transform method;then,referring to the masking threshold calculated by the psychoacoustic model,the "uniform quantization" and "non-uniform quantization" embedding algorithms are proposed.The "uniform quantization" algorithm spreads the quantization noise to each frequency point,thereby improving the imperceptibility of the algorithm;the "non-uniform quantization"algorithm uses the quantization process of the MP3 encoder and uses non-uniform quantization to improve the signal-to-noise ratio of the small signal,thereby improving the robustness of the algorithm.The experimental results show that the coding algorithm proposed in this paper not only has good invisibility,but also can effectively resist various noise attacks,and has excellent robustness.3.The paper proposed a differential decoding algorithm based on spread spectrum.The decoding algorithm combines the spread spectrum technology to calculate the hidden information by calculating the correlation between the audio information and the spread sequence.At the same time,the algorithm introduces a linear function to the two key factors affecting the decoding performance:original signal interference and noise interference.Eliminate the original signal interference,and use the differential method to reduce the impact of noise.And,the algorithm analyzes the impact of different decoding frequencies on decoding.Experimental results show that the decoding algorithm proposed in this paper can effectively resist noise attacks.