Steganalysis of 3D objects using statistics of local feature sets

Abstract

3D steganalysis aims to identify subtle invisible changes produced in graphical objects through digital watermarking or steganography. Sets of statistical representations of 3D features, extracted from both cover and stego 3D mesh objects, are used as inputs into machine learning classifiers in order to decide whether any information was hidden in the given graphical object. The features proposed in this paper include those representing the local object curvature, vertex normals, the local geometry representation in the spherical coordinate system. The effectiveness of these features is tested in various combinations with other features used for 3D steganalysis. The relevance of each feature for 3D steganalysis is assessed using the Pearson correlation coefficient. Six different 3D watermarking and steganographic methods are used for creating the stego-objects used in the evaluation study.