An Adaptive Steganography for 3D Triangle Models

摘要

In this paper, a new adaptive, highcapacity steganography for 3D triangle models using multi QIM is proposed. Even vertex of a 3D model can adaptively embed variable 3σ(σ≥1) bits using contour space subdivision with low distortion. We first use normal direction of vertexes to estimate the embedding capacity of even-vertex with respect to human visual system, then contour space subdivision is used to project the vertex to contour plane, finally at least three bits can be embedded by multi QIM in a vertex. The experiment results show that the proposed technique is adaptive, has high capacity' and low distortion, and is robust against afflne transformations and vertex reordering attack.