Optical image encryption with high efficiency based on variable-distance ghost imaging

In this paper, we propose an efficiency-enhanced optical image encryption system based on compressive computational ghost imaging (CCGI) with variable distance. Compressive sensing (CS) is known to be an effective algorithm to reduce the required number of measurements for a good reconstruction quality in cryptosystems based on computational ghost imaging (CGI). However, thousands of two-dimensional (2-D) random phase masks (RPMs) are still required as keys for one decrypted image in the conventional CCGI-based cryptosystems. This results in low efficiency of key transmission or storage. Hence, a novel optical cryptosystem based on variable-distance CCGI is proposed. In the proposed system, the distance from the RPM to the object can be shifted, and the same number of measurements can be realized by much less RPMs with various distances. It has been demonstrated that good reconstruction quality can be realized by only one RPM with various distances using the proposed approach. Compared to the conventional CCGI-based cryptosystems, many 2-D RPMs are saved for the same decryption quality in the proposed system, with a one-dimensional (1-D) vector consisting of various distances as a key instead. It is shown that the burden of key transmission or storage in the proposed system is reduced significantly. Meanwhile, the influence of the shifted distances on the feasibility, key sensitivity and the robustness against noise attack of the proposed cryptosystem has been investigated by simulation experiments.