Cryptanalysis and security enhancement of optical cryptography based on computational ghost imaging

Abstract

Optical cryptography based on computational ghost imaging (CGI) has attracted much attention of researchers because it encrypts plaintext into a random intensity vector rather than complexed-valued function. This promising feature of the CGI-based cryptography reduces the amount of data to be transmitted and stored and therefore brings convenience in practice. However, we find that this cryptography is vulnerable to chosen-plaintext attack because of the linear relationship between the input and output of the encryption system, and three feasible strategies are proposed to break it in this paper. Even though a large number of plaintexts need to be chosen in these attack methods, it means that this cryptography still exists security risks. To avoid these attacks, a security enhancement method utilizing an invertible matrix modulation is further discussed and the feasibility is verified by numerical simulations.