Research Of Audio Steganography Technology Based On Wavelet Transform

The development of computers and the Internet has greatly facilitated the transmission and sharing of network resources.However,the increasingly serious problems such as privacy leaks have brought severe challenges to network security.Although the traditional encryption technology can convert the intelligible plaintext to unreadable cipher text and protect secret information,it also exposes the importance of secret information,which can easily arouse the curiosity of third parties.Steganography,as an important component of information hiding technology,has become an effective complement to encryption technology.Unlike encryption,steganography is to hide secret information to other unimportant carrier,which is equivalent to putting on an “invisibility cloak” for secret information and makes the third party not aware of the existence of secret information,reducing the possibility of being found during the transmission of secret information.Because of its wide source and large redundancy,audio has broad application prospects in steganography.Therefore,the study of audio steganography has important theoretical value and practical significance.In this paper,the research object is audio steganography technology based on wavelet transform and three audio steganography algorithms are proposed to improve the performance of audio steganography.The main work and academic achievements are as follows:1.We deeply study the current audio steganography algorithms and analyze its advantages and disadvantages,then briefly introduce the general model of audio steganography and related basic concepts and expound the characteristics of human auditory system and audio digital features used in audio steganography,and then give the evaluation standard of audio steganography performance.2.In order to meet the needs of transmitting more secret information,an audio steganography algorithm based on integer wavelet transform and run length encoding is proposed.The algorithm improves the steganography capacity by using the lossless compression characteristics of run length encoding.Secret information is embedded into the wavelet domain coefficients.First,perform integer wavelet transform on the original audio and run length encoding on secret information,then get the secret data that needs to be embedded—secret data value and run length.Second,by setting a threshold,the secret data values are embedded into second bit and third bit of the integer wavelet coefficients whose absolute values are greater than the threshold.Finally,if the integer wavelet coefficients after being embedded the secret data are less than the threshold,the coefficients will be adjusted to make them greater than the threshold value via fine-tuning without affecting the secret data,which will ensure that the embedding operation does not cause the perception of the human auditory system and the successful extraction of secret information.The experimental results show that the capacity the proposed algorithm is greatly improved while ensuring the quality of stego audio.Compared with the existing methods,the steganography capacity of the algorithm has obvious advantages.3.In order to improve the quality of stego audio and ensure the security during transmission,we propose an adaptive audio steganography algorithm based on wavelet packet energy.The main energy of the audio is concentrated in the low frequency approximation coefficient,and the detail coefficients in the middle and high frequency are mostly the edges of the signal,so the change of the high frequency energy will have little effect on the original audio.In this algorithm,first calculate the wavelet packet energy of each wavelet packet sub-band;then we introduce the concept of weighted energy concentration and a threshold,adaptively select the embedding position and control the wavelet packet for embedding in the middle and high frequency,so as to improve the imperceptibility of the algorithm.Finally,four wavelet packets above the threshold are selected as embedding positions,and the secret information is embedded by adjusting the relationship among the four wavelet packets.Experiments show that the quality of the two test stego audio is very close to the highest quality audio test value.4.When stego audio is transmitted over a poor channel,it is subject to external interference.In order to improve the anti-interference ability of stego audio,centroid-based audio steganography algorithm in wavelet domain is proposed.In this algorithm,the concept of centroid is extended from frequency domain to wavelet domain,and the centroid of wavelet domain is considered as an influencing factor of embedded position.In this algorithm,secret information is represented by the XOR value of three fixed bits and one random bit.The random bit is determined by the centroid of each original audio segment.The wavelet domain centroid has good stability,and the secret information is expressed by the XOR value of four bits,which improves the fault tolerance of the algorithm.The experimental results show that this algorithm exhibits better robustness in resisting some common audio attacks.Compared with other methods,this algorithm has no disadvantage in steganographic capacity and imperceptibility.