LSB-based Random Embedding Image Steganography Technique Using Modified Collatz Conjecture
Aimee D. Molato, Fredilyn B. Calanda, Ariel M. Sison, Ruji P. Medina
Abstract
With digital communication breakthroughs, reliable data security and secure sharing of secret information remain a challenge. LSB is considered one of the most used techniques in steganography because of its simple and easy implementation. However, the algorithm can easily be broken down, and the attacker can obtain the secret message due to its simplicity. This paper combines Diffie-Hellman Key Exchange with a modified Collatz Conjecture to generate unique random numbers. The generated random numbers are used to identify the unique pixel locations where the secret message will be embedded. The simulation results using INRIA Holiday datasets yielded PSNR and SSIM mean values of 69.24 and 0.9999, respectively, which are favorable enough to conclude that the employed method generates highly imperceptible stego-images. Furthermore, the probability of detection using statistical attacks like RS Analysis and Sample Pair Analysis was fortunately minimal, leading to stego-images that are resistant to statistical attacks.