Light-field 3D encryption based on the monocular depth rendering

Abstract

The real-time performance of light-field 3D encryption technology based on the integral imaging principle is restricted by the acquisition speed and the data of the elemental image array (EIA). Herein, we propose a light-field 3D encryption scheme based on monocular depth rendering. With the help of a convolution residuals network (CRN), the proposed scheme can generate the corresponding depth map from a single RGB image and simplify the pickup process of the EIA according to the image mapping. For encryption, using reversible state loop cellular automata (RSL-CA) to encrypt a single RGB image updates traditional 3D encryption, greatly improving the security and efficiency of the encryption algorithm. It is experimentally demonstrated that optical 3D reconstruction is clear and brightly colorful and also has a good parallax effect. The proposed method can open a brand-new research perspective for light-field 3D encryption.