SNAP-compressive lossless sensitive image authentication and protection scheme based on Genus-2 hyper elliptic curve

Abstract

Hyper Elliptic Curve Cryptography (HECC) offers comparable security as that of Elliptic Curve Cryptography (ECC) even with much reduced key size. This reduction in key size minimizes processing overhead and hence is particularly more attractive for the constrained environment, key agreement protocols, and digital signature applications. In the digital communication world, the most dominant communication happens in Image form, which is highly sensitive as they possess biometric information or individual’s privacy information like medical history. As images consume large storage, it needs to be compressed for faster processing before encryption without compromising security. This, of course, incurs additional overhead at the receiver end along with data loss. To minimize storage space by compression without any loss of data and to speed up the encryption process without compromising security a lightweight model named SNAP (SeNsitive image Authentication Protection) has been proposed. This SNAP employs genus-2 HECC that can be constructed on 80-bit finite field to attain security features as compared with 160 bit ECC or 1024 bit RSA. Since HECC comes under Discrete Logarithm Problem (DLP) based cryptography, solving the DLP with the prime field of size 256 bits is not feasible with modern quantum computing power. It compresses the points of the chosen curve to divisor by using Mumford representation. This representation can perform compression without any loss which is desirable for ensuring sensitive protection. As HECC processing relies on divisors rather than points, it is faster than EC based image encryption. Rigorous security analysis has been conducted on the proposed SNAP model. The observed results claim ideal measures for MSE and PSNR values in comparison against a contemporary benchmark image encryption scheme. Security strength of the SNAP method has been ratified by examining key space analysis, entropy analysis, NPCR, and UACI analysis. The proposed method can be recommended as a good candidature as it out-performs in all the above performance measures from the security perspective.