Data hiding in compressed images

Abstract

Apart from the obvious ethical issues, digital information manipulation poses security issues. Ensuring an omnipresent balance amongst the security triad of confidentiality, integrity and especially availability is not a trivial feat. More than often compromises need to be done which risk jeopardizing digital content. This dissertation brings together work done by various senior researchers in the last ten years in the fields of data hiding, authentication and tampering localization. Various facets, such as reversibility and fragility are reviewed. Though most of the algorithms relate to the spatial domain, a few appertain to other domains, through the use of frequency and wavelet transforms. Several algorithms are actually implemented as proposed by their originators whilst others are incorporated as the basic processor on which a higher level algorithm can be built. A graphical user interface was also developed to demonstrate the resulting transformations and to enable the analyses and comparison of certain widely agreed metrics.