Medical Color Image Encryption Using Chaotic Framework and AES Through Poisson Regression Model

Abstract

This paper suggests a novel diversion in color medical image encryption using a chaotic framework and Advanced Encryption Standard AES with Poisson regression model. Nowa-days, the remote healthcare monitoring application is getting prominent by providing better assistance to people's life. We proposed a secure color image encryption algorithm for the medical images using the 2D Arnold cat map, AES-128 and Poisson regression. The workflow explained sequentially in this way. First, the plain medical image is decoupled into the corresponding RGB channels. Next, the chaotic map is applied to the plain image for converting it into a scrambled one. This scrambled image is transmitted to the AES-128 encryption block which converts the scrambled image into the encoded text form and encrypted using the hashed symmetric Key. Then the Encrypted image is formed through the Poisson regression model to predict the pixels based on the text encrypted. Finally, the resultant image is transmitted to the receiver with the NPCR score of 99.0174 and average UACI score of 33.0690. The results for the experimental work and its formulated security analyses reveal that this image encryption technique is applicable for medical image encryption and transmission.