Data hiding on point‐sampled geometry

Abstract

We present two new schemes, SSA and ESA, for digital steganography of pointsampled geometry in the spatial domain. Both schemes employ principal component analysis (PCA). The SSA approach increases data hiding capacity by nearly 1.5 times the original capacity. ESA achieves between 1.2 and 3 times the model capacity. Experimental results verify that our schemes can achieve high data capacities with insignificant visual distortion of the original model.

Key words:

• PCA
• point‐sampled geometry
• steganography

Notes

Corresponding author. (Email: [email protected])