UDC 004.422.832
Authors
Kustov V. N.
Saint Petersburg state University of railway transport of Emperor Alexander I, Informatic and Information
Security Department Saint Petersburg, Russia
Krasnov A. G.
JSC "Nexign", Saint-Petersburg, Russia
Abstract
This article presents a software model of a highly undetectable stegosystem that uses digital still images as covering objects. The main principle of using the model is to minimize adequately specified distortion using a new efficient encoding algorithm. Distortion is defined as an operation ⊕ (excluding OR) on a specially selected byte of the image (covering object), which is selected from the condition of minimizing the penalty function calculated on some selected neighborhood of this byte. This allows authors to keep the original image almost unchanged and make it indistinguishable even after large payloads are embedded in it. Thus, the embedded hidden message may not be visible when the steganalysis even for the experienced steganalytic. The stegosystem model is called the ⊕Highly Undetectable SteGOsystem model. In short, it can be referred to as the ⊕HUGO model. This model can be implemented to embed hidden messages in the spatial area of digital still images much more efficiently than using the LSB algorithm. The proposed stegosystem model provides higher performance than the LSB algorithm and allows you to transmit a longer message, while providing higher resistance to detection.
Keywords
Steganography, Stegosystem, Software model, Highly undetectable stegosystem, ⊕HUGO
