Adaptive image steganography based on transform domain via genetic algorithm

This paper presents a novel approach for data hiding in frequency domain with the use of genetic algorithm. At first, cover images are mapped to a proper frequency domain using the concepts of adaptive wavelet transform and genetic algorithm. In the obtained space, using a model based on Kieu and Chang, encrypted information will be embedded in the frequency coefficients that represent edges of the image in spatial domain. So the cover image will change the least and have the most compatibility with human visual system. Simulation results show that our proposed method outperforms recently published works in terms of PSNR and PSPNR factors.