Anti-Forensics with Steganographic File Embedding in Digital Image Using Genetic Algorithm

Abstract

In this study, a steganography method on digital images as anti-forensics by utilizing genetic algorithms was proposed. Genetic Algorithms are artificial intelligence whose functions are optimization and search. The purpose of this research is to optimize steganography as anti-forensic by applying a Genetic Algorithm and combined with the Hilbert curve, lempel Ziv Markov chain, and least significant bit. The result provides a new steganography method by combining various existing methods. The proposed method will be tested for image quality using PSNR, SSIM, Chi-Squared steganalysis and RS-Analysis, and extraction test. The novelty obtained from the developed method is that the steganography method is as optimal as anti-forensic in keeping confidential data, has a large embedding capacity, and is able to be undetected using forensic methods. The results can maintain data confidentiality, have a large embedding capacity, and are able to be undetected using forensic methods. The proposed method got better performance rather than the previous method because PSNR and SSIM values are high, secret data can be received back as long as the pixel value doesn't change, and the size of the embedding capacity. The proposed method has more ability to embed various types of payload/ secret data because of the way it works, which splits byte files into binary. The proposed method also has the ability not to be detected when forensic image testing is carried out.