标题：Two-level image authentication by two-step phase-shifting interferometry and compressive sensing
作者：Zhang, Xue; Meng, Xiangfeng; Yin, Yongkai; Yang, Xiulun; Wang, Yurong; Li, Xianye; Peng, Xiang; He, Wenqi; Dong, Guoyan; Chen, Hongy 更多 作者机构：[Zhang, Xue; Meng, Xiangfeng; Yin, Yongkai; Yang, Xiulun; Wang, Yurong; Li, Xianye; He, Wenqi] Shandong Univ, Sch Informat Sci & Engn, Dept Opt, Jinan 更多
通讯作者地址：[Meng, XF]Shandong Univ, Sch Informat Sci & Engn, Dept Opt, Jinan 250100, Shandong, Peoples R China;[Meng, XF]Shandong Univ, Shandong Prov Key Lab Las 更多
来源：OPTICS AND LASERS IN ENGINEERING
关键词：Compressive sensing; Two-step phase-shifting interferometry; Image; encryption; Image authentication
摘要：A two-level image authentication method is proposed; the method is based on two-step phase-shifting interferometry, double random phase encoding, and compressive sensing (CS) theory, by which the certification image can be encoded into two interferograms. Through discrete wavelet transform (DWT), sparseness processing, Arnold transform, and data compression, two compressed signals can be generated and delivered to two different participants of the authentication system. Only the participant who possesses the first compressed signal attempts to pass the low-level authentication. The application of Orthogonal Match Pursuit CS algorithm reconstruction, inverse Arnold transform, inverse DWT, two-step phase-shifting wavefront reconstruction, and inverse Fresnel transform can result in the output of a remarkable peak in the central location of the nonlinear correlation coefficient distributions of the recovered image and the standard certification image. Then, the other participant, who possesses the second compressed signal, is authorized to carry out the high-level authentication. Therefore, both compressed signals are collected to reconstruct the original meaningful certification image with a high correlation coefficient. Theoretical analysis and numerical simulations verify the feasibility of the proposed method. (C) 2017 Elsevier Ltd. All rights reserved.