Steganographic methods in vector graphics

Abstract

Various steganographic techniques are used to hide information. Usually, information is hidden in images, audio and video files, text documents, and the like. The article deals with vector images consisting of various mathematical objects (points, lines, curves of the first and second order, Bezier curves, nodes, tangents, base points, etc.). Information hiding techniques alter these mathematical objects, for example, by encoding the coordinates of the base points. The most successful for carrying out steganographic transformations is the SVG vector graphics format, which, due to its structure, makes it easy to manipulate the objects of which it consists. Its broad support across platforms also allows for increased secrecy when transferring sensitive data by sending seemingly ordinary media files. The article discusses two methods (bitwise and the method of patterns) of hiding information in vector images, studied their features, advantages and disadvantages. Various affine transformations that can be used to disrupt the operation of the steganosystem were also investigated. The most common types of affine transformations are the operations of transfer, rotation, shift and scaling with possible variations (offsets along the abscissa and ordinate axes, proportional and non-proportional scaling, with compression and expansion). Most of the methods for embedding information into vector images provide a one-time resistance to affine transformations, while the repeated imposition of operations for changing the position of objects may destroy the message altogether. The methods investigated in the work (bitwise and the method of patterns) implement a higher level of resistance to various kinds of transformations when they are repeated many times, and the conducted experiments clearly demonstrate this. The results obtained show that vector images can indeed be used to hide information, but the resistance against certain affine attacks is not always high.