An Audio Steganography Method Based On Discrete Wavelet Transform And Singular Value Decomposition

In the field of information security,traditional encrypted communication is widely used,but there are two fatal weaknesses.First,the ability to crack complex encryption technology is enhanced,and second,the encrypted data is easy to attract the attention of attackers.Therefore,information steganography technology has been highly valued and deeply studied.Steganography can protect secret data while hiding the existence of secret data communication,so that non-communication recipients are not aware of the occurrence of secret data communication,and it is difficult to intercept secret data.In order to make the audio steganography algorithm universal,this thesis designs an audio steganography algorithm that can improve the hidden capacity and imperceptibility of the algorithm,and the main work content is as follows:Firstly,in the preprocessing part of the audio steganography algorithm,in order to improve the security of the algorithm,the encryption of the target information is completed using two logistic chaos sequences.At the same time,the algorithm chooses to fine-tune and perturb the logistic chaotic sequence,which solves the possible periodic problem of the logistic chaotic sequence.According to the analysis and verification of experimental results,it is proved that this choice can better realize the encryption and decryption of secret audio under logistic chaotic sequence.Secondly,an improved audio steganography algorithm for discrete wavelet transform and singular value decomposition is proposed.The original audio datas are transformed by discrete wavelets to obtain the detail coefficients and the approximate coefficients.The traditional audio steganography algorithm embeds the datas related to the secret audio into the detail coefficients,and the algorithm embeds them into the key obtained after the approximate coefficients are decomposed by the singular value,which effectively improves the imperceptibility of the audio.The setting of double chaos encryption also improves the security of the algorithm to a certain extent.In addition,the change of secret datas embedding strength and embedding position further improves the hidden capacity of the algorithm.When the secret datas are embedded at the 100,000 th position point of the original audio datas at an embedding intensity of 0.008,the hidden capacity can reach up to19110 bps.Finally,the improved audio steganography algorithm is experimentally verified and analyzed.The algorithm performance is verified by changing the three factors of embedding strength,embedding position and variable amount of transmission medium.Experiments show that the improved audio steganography algorithm has large hidden capacity and better imperceptibility.At the same time,the participation of the initial carrier audio is not required in the extraction process of the secret audio,and the blind extraction of the secret audio is realized.