Image authentication using genetic algorithm

Abstract

With a widespread deployment of digital images, there is need to authenticatesuch images for their originality and integrity.The main objective of the work is todesign an authentication scheme that allows two parties to exchange images witheach other while ensuring both integrity and authentication. In other words, wewant an authentication scheme in which any content modification on the imagewill be considered unacceptable . In this work, an image authentication schemebased on the Genetic Algorithm is proposed. The scheme uses Crossover andMutation methods of Genetic Algorithm to produce non-reversible compressionof images into random binary strings with the use of secret key to be used onlybetween the sender and the receiver.In the proposed scheme, the original image is employed as the initial populationfor starting the genetic algorithm, the employed initial population is then appliedto genetic algorithm in which the result obtained from previous iteration is iteratedagain and again so that the best optimized image with lowest correlationcoefficient among the adjacent pixels is produced. Finally a modulus operator isused which converts the variable length message to a fixed length output(digest)that can be appended with the message to be sent at the receiver end.The samedigest can be calculated at the receiver end to ensure the originality of image.The primary requirements of the hash function that is pre-image resistance, collisionresistance has been proved and the running time performance of the schemeis compared with other standard cryptographic authentication schemes such asmessage authentication cod