Abstract
Steganography is the process of concealing data to protect the carrier file data from intruders. One of the main challenges of steganography is to maintain optimum visual quality while increasing hiding capacity. This can be resolved by incorporating the human visual systems weaknesses in steganography, which results in a good quality of the algorithm. Additionally, steganography is considered as an optimization problem to obtain optimum embedding rate. For this reason, this paper presents a novel method aimed towards a selection of perceptual embedding threshold in Discrete Wavelet Transform using human visual system characteristics and Genetic Algorithm. This method included an introduction of an optimization model by maintaining a correlation between neighboring areas of the image and the different parts of the object. The results of the experiment involving a variety of two thousand images show the differences between the various methods and the HVS in relation to accuracy. Furthermore, higher levels of detectability and comprehensiveness were seen. Lastly, the method has a high tendency to maintain image quality as well as the transparency when subjected to steganography.