Multiple images simultaneous encryption and decryption via deep-learning assisted interferenceless coded aperture correlation holography

Abstract

Optical encryption based on interferenceless coding aperture correlation holography (I-COACH) is one of the holographic encryption technologies. I-COACH with dual-channel has been employed to encrypt two plaintexts, not only is the encryption capacity limited, but when the two plaintexts are horizontally overlapped, the crosstalk between the two decrypted plaintexts becomes significant. To address these issues, we present an optical multiple images encryption system based on single-channel I-COACH. In our optical encryption system, the multiple images are placed adjacent to each other and encrypted simultaneously through a single-shot I-COACH. The traditional cross-correlation methods fail to decrypt the multiple plaintexts, we propose a decryption scheme based on a novel deep learning framework (M-Net). The multiple plaintexts can be decrypted from a single ciphertext simultaneously, greatly improving the decryption efficiency by our method. Moreover, the crosstalk between the multiple images is avoided. Various experiments are conducted to demonstrate the effectiveness and feasibility of our proposal.