.

ISSN 2063-5346
For urgent queries please contact : +918130348310

DEEP LEARNING BASED IMAGE ENCRYPTION AND DECRYPTION

Main Article Content

DEEP LEARNING BASED IMAGE ENCRYPTION AND DECRYPTION
» doi: 10.31838/ecb/2023.12.s1-B.393

Abstract

The suggested approach generates a sample medical picture. A new hybrid security algorithm for the RSA cryptosystem is described. For decryption and encryption, the system employs two distinct keys: a private key and a public key. As a result, it provides a more secure channel for the encryption and decryption processes. The phi generates the value of n public keys and n private keys, as well as completing the decryption and encryption operations. A reconstruction network is used to restore the encrypted image to the original (plaintext) image. To ease data mining straight from the privacy-protected environment, an area of interest (ROI)-miningnetwork is developed to extract the interesting item from the encrypted picture. The suggested method is evaluated using a medical X-ray dataset. Comprehensive experimental results and security evaluations show that the proposed approach may give a significant level of safety while remaining efficient. This project's front end is written in MATLAB JAVA. The identification of medical photographs is widespread. To encrypt and decrypt the binary data, we employ the RSA model for safe encryption and the stegno image model. The most famous and commonly used cryptosystem is RSA, whose security is determined by the difficulty of obtaining the private key in an acceptable amount of time rather than the algorithm's specifics. RSA requires the highest type of representation in the picture encryption field to achieve maximum security

Article Details