Image encryption based on moiré pattern performed by computational algorithms

A technique based on optical operations for image encryption and decryption is presented by simulation and experimentally. In this method, an image is encrypted by a fringe pattern. This fringe pattern is generated by a computational algorithm as a cosine function, which added in its argument the intensity image as a reflectance map. The result of the encryption process is a fringe pattern deformed according to the image reflectance map. The decryption method is performed creating a moire fringe pattern. To carry it out, the encrypted image is overlapped with a key fringe pattern. This key code is an undeformed fringe pattern, which is generated at the same frequency of the encrypted image. The obtained moire pattern is a modulation function, whose envelope corresponds to an approximated version of the original image. To extract the envelope, a low pass filter is applied on the moire pattern. This encryption technique and decryption is applied to the face images. To describe the accuracy of this technique, a root mean square (rms) of error is calculated using data from the decrypted image and data from the original image. This technique has a great potential, because the process is performed using computational algorithms based on optical operations. In this manner, optical components are avoided. It is an advantage over optical methods for encryption and decryption, which use optical components. This technique is tested with simulated images and real images and its experimental results are presented.