On the Efficiency of Metaheuristic Optimization for Adaptive Image Steganography in the DFT Domain

Abstract

Adaptive embedding is a popular area of image steganography. Adaptability means the use of cover image features in the embedding process. In many cases, the embedding adaptability is associated with choosing the best position of the message bits in the image. However, the total check of all possible options for the message fragment is redundant in most cases. This leads us to the optimization problem. In this paper, we investigate the applicability of some classical metaheuristic optimization algorithms for solving the problem of increasing the efficiency of adaptive embedding of information into the discrete Fourier transform phase spectrum. We consider three popular optimization algorithms: the genetic algorithm, the differential evolution algorithm, and the particle swarm optimization algorithm. The experimental results show that the particle swarm optimization algorithm can significantly increase the embedding capacity, while the imperceptibility also improves or remains at the same level, and the embedded information is extracted without any errors.