Optical phase image encryption using stokes parameters and singular value decomposition

In this paper, we propose an optical asymmetric phase image encryption method in which the vectorial light field is used to encode the data. In transverse plane, the vectorial light field has spatially varying polarization distributions where we are allowed to have a greater number of degrees of freedom. In this scheme, the input image is first phase encoded and then modulated by a phase encrypting key, synthesized from the speckles obtained by the scattering of Hermite-Gaussian beams. The modulated image is further processed using fractional Fourier transform with a specific order (α). A pixel scrambling operator is utilized to increase the randomness to further enhance the security and singular value decomposition approach is employed to add the nonlinearity in the encryption process. Now, the stokes parameters, i.e. S₁ and S₂ are calculated using the light intensities correspond to different polarizations. S₁ is used as the encrypted image for transmission and S₂ is reserved as one of the private decryption keys. The robustness of the proposed technique is tested against various existing attacks, such as known plaintext attack, chosen plaintext attack, and contamination attacks. Numerically simulated results validate the effectiveness and efficiency of the proposed method.