Multylevel Audio Structure for Weighted Information Embedding

Abstract

This article presents the research dedicated to the problem of possibility to use weighed embedding approach to increase the efficiency of audio watermarking. It was shown that it is possible to highlight at least three non-equal (with different influence of embedded errors) zones for DCT (discrete cosine transform) coefficient of audio mono-signal. This result shoes that it is possible to construct weighted structure model for audio signal that is corresponds with a Human Auditory System Model.