Image encryption based on phase encoding by means of a fringe pattern and computational algorithms

Abstract

A computational technique for image encryption and decryption is presented. The technique is based on light reflection, intensity superposition and computational algorithms. The image to be encrypted is a reflectance map obtained by means of the light reflected by a scene. To perform the encryption procedure, the image is encoded in a computer-generated fringe pattern. The model of the fringe pattern is a cosine function, which adds to its argument the image to be encrypted as a phase. It generates a fringe pattern deformed according to the image. To complete the encryption, a random mask is superimposed on the fringe pattern. The decryption procedure is performed by subtracting the random mask from the encrypted image and applying a phase recovery method. To retrieve the phase from the fringe pattern, the heterodyne demodulation method is used. To describe the accuracy of results of the decrypted images and the robustness of the encryption, a root mean square of error is calculated. All steps of the encryption and decryption are performed in computational form. The results of encryption and decryption are thus improved. It represents a contribution to the field of encryption and decryption. This technique is tested with simulated images and real images, and its results are presented.