Variant of LSB Steganography Algorithm for Hiding Information in RGB Images

Abstract

© 2017 SERSC. Steganography has been emerged as a prominent technique for communicating secret information in a proper multimedia career. In image steganography, least significant bit (LSB) substitution method is popular but vulnerable to numerous security attacks. Hence, it is always imperative to make the steganographic technique more robust and secure. This paper presents a new approach to hiding text messages within red-green-blue (RGB) images based on LSB technique. Unlike the standard LSB technique, where message bits are directly embedded in the LSB positions, our proposed algorithm encodes these message bits before embedding in the LSB positions. This embedding and encoding process are done on the basis of most significant bit (MSB) values of RGB and the concept of odd and even parities of those pixels respectively. The main goal of the presented method is to improve the robustness and the image quality of the stego-image. The performance of stego-image quality is measured by mean square of error (MSE), peak signal-to-noise ratio (PSNR), structural similarity measure (SSIM), bits ratio (BR), and histogram analysis. Experimental results and comparative studies have revealed that the proposed technique is able to conceal message effectively, with there being little difference between the cover image and the stego-image. Results also show that our proposed approach can embed reasonable amount of information.