A blind steganalysis scheme using estimation techniques

Distributed image steganography (DIS) is a method for hiding a secret image inside n different innocuous cover images. The method is secure and reliable because it leaves very little traces of the secret image in the stego image. However, it is possible that criminals could use it for unchallenged, covert communication of illegal secret information over the Internet. In such cases, a suitable countermeasure (also known as steganalysis), to reverse the DIS is needed which can detect and disclose this kind of secret information. In this paper, the challenge of disclosing the secret image was handled by adopting an estimation theoretic approach using techniques such as Kalman filtering. Another unique feature is the blind steganalysis approach which assumes no knowledge of the cover images. The results indicate the potential for its usage by law enforcement agencies in detecting and extracting secret information hidden by the DIS.