A Comprehensive Study on the Homomorphic Encryption for Secure Image Data Processing

In the contemporary digital landscape, the integrity and confidentiality of data have become paramount concerns. This study presents a comprehensive framework for secure image data processing using homomorphic encryption. The proposed approach involves image preprocessing, logistic regression model training, feature extraction, and polynomial approximation to accommodate the constraints of homomorphic encryption algorithms. Sensitive data, encrypted via homomorphic algorithms, is embedded within images to ensure its concealment during computational operations. Subsequent encryption of the image using the asymmetric Rivest-Shamir-Adleman (RSA) algorithm further secures the encapsulated sensitive data. Through experimental data and analysis, the performance and speed of homomorphic encryption are compared against traditional methods, validating its efficacy in encrypted image data processing.