Dual Image-Based Dictionary Encoded Data Hiding in Spatial Domain

Abstract

In the modern digital era, the privacy of personal communication is a serious concern to all netizens. A better way to preserve privacy could be to hide the secret message inside some innocent looking digital object such as image, audio, video, etc., which is known as steganography. A new steganographic scheme using a reference image along with the cover image has been proposed in this article. It enhances the robustness and security by increasing the obscurity of the hidden message. It also employs an additional dictionary-based encoding module to increase the hiding capacity as well as security. Experiments show that bit changes in the reference image are very few and undetectable to human perception, it also evades common statistical tests. Evaluation of standard quality parameters such as MSE, PSNR, UIQI, SSIM along with chi-squared statistics based embedding probability testing has been performed. When dictionary-based encoding is applied it further improves the quality parameters.