On the stability of subband hidden embedding data into images based on cosine transform subband matrices

Abstract

We considered one of the approaches to solving the urgent problem of copyright protection for images stored in modern information and telecommunications networks, based on the hidden embedding of the author identifying information directly into the protected files in this article. The main provisions of the method of sub-band hidden data embedding in images based on the relative change of the container image projections on the cosine transform sub-band matrices eigenvectors were presented. On the basis of computational experiments, the resistance to external destructive effects in the form of additive random noise of data embedded on the basis of this method was investigated. When conducting the computational experiments, we used three images containing a different number of large and small objects. The embedded data was presented in bit form. Additive random noise was generated using a Matlab function. The results of computational experiments have shown that the method of sub-band hidden embedding has an advantage over the considered known methods of hidden embedding from the standpoint of the embedded data recovery stability to the effects of additive random noise.