About the method for hidden coding of control information to speech data

Currently, the industry's creation of information, educational and entertainment content for audio information inquiries, movies and music are widely used spoken language. In this connection there is the problem of providing automatic control of the use of speech and in particular to prevent unauthorized actions with it. In many ways, this control is advantageously carried out in a secretive mode, where information about the processes of encoding and appropriate action available only to a specific group of persons. The authors have to solve this problem, we propose a method of adaptive hidden coding control information, providing for a given error probability of high secrecy. The method is to use the power properties of the voice data, the mathematical basis is used as an orthogonal basis of eigenvectors sub-band matrix instead of a pseudo-random sequence (PRS), which is widely used at present in the hidden coding control information. The studies found that the proposed method of sub-band projections has a higher level of secrecy of information control over the method of spreading since it causes less change in the proportion of the energy of speech segment data. Minimizing variations in energy band of the portion where the coding is not carried out, it is achieved through the use of sub-band projections found for sub-band matrix of eigenvectors whose eigenvalues are close to unity.