Research On Technology Of Audio Data Hiding

Audio information hiding is an important technology in the field of information security.In early time,audio information hiding was used for digital audio source copyright protection,but nowadays a variety of application scenarios have been developed,such as hidden introduction information in scenic broadcast audio,hidden shopping information in film and television works,and so on.Existing audio information hiding technologies are increasingly unable to meet the needs of these scenarios,and further research on this technology has also been put on the agenda.The research in this paper is mainly aimed at frequency domain spread spectrum audio information hiding system,which solves two problems in this system:First problem,how to suppress the interference of host signals;Second problem,how to achieve reliable and effective information receiving in low-SNR-noise environment.To solve the first problem,we proposed two kinds of algorithms suitable for spread-spectrum systems in frequency domain based on the psychoacoustic masking model:1.One is an information hiding algorithm based on differential frame structure,which uses the short-term stationarity of the signal to suppress host signal interference.This algorithm corresponds to two kinds of information extraction methods,and directly performs difference and logarithm difference.Both information extraction methods have their own performance in various SNR noise environments.2.The other is an information hiding algorithm based on correlation compensation.When embedding information,an appropriate compensation component and high-pass function are designed.This compensation component and high-pass function perfectly cancel the host signal component when extracting information,eliminating the host.Signal interference.Experiments show that this method is robust against digital attacks such as resampling and compression.To solve the second problem,this paper also proposes a despreading method based on noise estimation and the corresponding synchronous decoding system design,which is used to deal with the noise environment with low SNR.The despreading method of noise estimation can be used for despreading of the synchronization process and the decoding process.We use the spectral distribution characteristics of the audio signal to establish a reception threshold and improve the reception ratio of the effective frequency.Experiments show that the synchronization efficiency and decoding efficiency of the noise estimation technique are higher than the traditional despreading method.