A zero-watermarking scheme for three-dimensional mesh models based on multi-features

Abstract

The aim of this study is to devise a zero-watermarking algorithm, the main challenge of which is the selection and construction of resilient features of three-dimensional(3D) models while improving the resilience of the zero-watermarking algorithm to common attacks. A novel zero-watermarking scheme for 3D mesh models based on multi-features is proposed in this paper. Using the Ordered Statistics Vertex Extraction and Tracing Algorithm, the stable vertices of a 3D model are determined. Then, the skewness measure of the shape diameter function values and vertex norms of the stable vertices are determined. Combined with the distribution feature of the number of stable vertices, a resilient watermark is applied. Experimental results show that the proposed scheme is not affected by a similarity transformation (such as translation, rotation, or uniform scaling) or vertex reordering. The scheme is also sufficiently resilient to noise, simplification, smoothing, and subdivision attacks. In addition, the scheme can resist cropping to a certain extent. The scheme proposed in this paper is suitable for applications that require 3D models with a higher accuracy and that cannot be embedded with a watermark that would modify the data of those models.