Research On Key Technique Of Information Hiding In WAV Files

With the rapid development of multimedia technology and communication network technology,the frequency of digital storage and transmission of information data has exploded,and the concerns about information security and intellectual property protection have also increased.With the advancement of technology,people in the traditional information security strategy,another way put forward based on information hiding technology information security and intellectual property protection measures.The technology in the audio data carrier also has a wide range of applications.But because of the physiological characteristics of the human ear is very sensitive to changes in the received audio signal,so if you want to embed audio data in the secret data in the image and other multimedia carriers embedded in cipher text data is much more difficult.After the cipher text information is embedded into the audio signal,the hidden signal will be disturbed by the channel noise during the transmission of the ordinary channel,and the data compression and decompression will be damaged during sending and receiving.The non-malicious synchronization and filtering will result in Data errors,the above-mentioned various adverse factors lead to the audio information hiding algorithm robustness and anti-attack have high requirements.The traditional information hiding method of audio signal mainly embeds cipher text information,minimum information bits,echo and other information hiding algorithms in the time domain is the most well-known information hiding algorithm.The information hiding algorithm based on time domain is characterized by simple implementation and large amount of hidden data.However,the shortcomings of the time domain algorithm is also very prominent,for example,the embedded part of the hidden vector and unmodified parts have different statistical properties,while the robustness of time domain algorithm is poor,resistance to geometric distortion,noise and data compression Even if the very simple array filter will interfere with the final cipher text information extraction.The main work of this paper is to study the characteristics of human auditory perception mechanism.Based on the human ear's auditory perception characteristics and the acoustic characteristics of audio signals,the signal processing technology of transform domain is used to hide the information in speech signals.The method of transform domain is simply to do discrete cosine transform on the carrier audio signal,or to obtain the MFCC parameter of the cepstral domain and do the wavelet transform on the carrier audio signal and embed the cipher text information into the carrier audio signal In the same time in the time domain requirements of the operation can not be perceived by the human ear.In summary,the following research results have been achieved:1)A speech information hiding method for embedding cipher text information after discrete cosine transform of carrier speech signal is proposed.The method mainly uses the discrete cosine transform operation characteristic,and embeds the ciphertext signal into the carrier signal.The embedding rule is:divide all the frame signals into high,medium and low according to the different energy values of each frame of the carrier audio signal,embed the different number of ciphertext information according to the energy of each frame signal energy,Low energy embedded or not embedded in the voice carrier signal itself to ensure the auditory characteristics of the premise,taking into account the embedded ciphertext information security.In the information extraction process,the data is classified and filtered according to the energy level of the signal energy by using the energy preserving and concentrating characteristic of the discrete cosine transform.After that,the low frequency coefficients are extracted from the difference signal by using the implicit signal frame and the original data frame To achieve ciphertext information and carrier information separation,thereby extracting the ciphertext information.According to the experimental results,this algorithm has high security and imperceptibility,and has some resistance to conventional channel noise attack and non-malicious low-pass filter attack.2)A speech information hiding method is proposed to hide secret messages in MFCC parameters of speech signals.This method mainly uses the strong stability of the short-term energy of the speech,changes the value of the MFCC parameter by a small amount to represent 0 or 1 bit of the secret information,and then reverses the calculated parameter by changing the parameter.The corresponding logarithmic spectrum is obtained.After indexing the logarithmic spectrum,a result is obtained,and then the MEL filter bank is used to filter the result to obtain a new linear spectrum;finally,the new linear spectrum is subjected to the inverse transform of FFT to obtain a new A time domain signal with hidden cipher text.This newly obtained implicit time-domain signal has little difference in auditive and original signal,so as to realize the concealment of the information.On the other hand,in the process of information extraction,we can extract the cipher text information from the implicit documents by using the frame synchronization during the concealment and extraction.This method is suitable for voice applications in local area network or high-speed network,and transmits secret information effectively on the basis of transmitting carrier voice without audible distortion.3)An audio information hiding algorithm based on wavelet domain is proposed.The algorithm uses the energy preserving and focusing feature of wavelet transform to transform the cipher text information into the scrambling sequence as the information to be hidden.Then the wavelet transform is used to decompose the carrier audio files quickly and easily.The cipher text information sequences after scrambling are embedded into the low frequency components of the carrier.Finally,the inverse wavelet transform is used to generate the hidden audio files.This algorithm can hide the cipher text information in the detailed component of the audio signal,that is,put the ciphertext information in the energy of the carrier audio-the low frequency part,which makes the algorithm anti-interference ability strong.Experiments show that the wavelet transform method can recognize hidden cipher text information and has good robustness for common compression,clipping and noise attack.The sound of the cipher text information can be anti-compression and anti-noise processing.This method is suitable for MP3 and other loss compression of multimedia files copyright protection.In this paper,the theory and application of speech information hiding are explored and researched,and some achievements have been made.However,the current technology of this research field is far away from the commercialization of large-scale applications and industrialization.There are still many problems to be solved,such as establishing and perfecting a set of objective and fair evaluation standard of hidden information.In the synchronous transmission technology,And the method of blind detection and extraction based on human auditory feature which can satisfy the commercial demand.Benefit from the continuous development of cryptography,multimedia technology,modern communication theory,coding theory,signal processing and pattern recognition and other interdisciplinary development,the study of phonological information hiding will also emerge new research ideas and methods.It is hoped that the research results of this paper can play a positive role in promoting the further development of information hiding technology.