IMPROVEMENT OF THE DIGITAL WATERMARK EMBEDDING ALGORITHM TO INCREASE ITS RESISTANCE TO GEOMETRIC TRANSFORMATION OPERATIONS IN THREE-DIMENSIONAL MODELS

Abstract

The development of computing power and data visualization methods has not only facilitated the widespread use of 3D models, but also increased the demand for them in various fields of human activity. This, in turn, has led to problems such as copyright infringement and disputes over the authenticity of the owner, which are common due to the ease of pirating 3D models. An effective solution to prevent these problems is the use of digital watermarks, which play a key role in protecting against illegal copying, modification, and falsification of various visual materials. In turn, algorithms for embedding and detecting digital watermarks in three-dimensional models must meet a number of requirements, including being resistant to various geometric transformation operations, thereby ensuring reliability and efficiency in the data processing process. In this regard, this paper proposes to improve one of the digital watermark embedding algorithms that uses the search for landmarks and is based on finding an estimate of the abrupt change in area. To achieve this goal, the first stage of the algorithm was modified by applying a 3D edge vertex recognition algorithm, which is more stable for watermark embedding and, accordingly, their search and detection are characterized by a high level of reliability. In addition, this algorithm has a number of advantages, in particular: it is quite flexible, as the support can be adapted both globally and locally; it can be used for any type of mesh, provided that an appropriate set of parameters is chosen; it is less affected by noise that is filtered out during the best fit process. In order to confirm the strengths of the proposed improved algorithm, we tested it on a complex model and proved that it is more resistant to a number of common geometric attacks on 3D mesh models and shows better results than the original algorithm for finding landmarks for embedding a digital watermark.