Optical Image Encryption and Classification Using Deep Neural Networks for IoT Applications
Poonam Yadav, Hukum Singh · 2024
In this paper, we present a novel method for encrypting images using double random phase encoding (DRPE) and securely storing them in the cloud without decryption. These encrypted images are classified using deep convolutional neural networks (CNN) in the cloud, eliminating the need for data decryption, and enhancing privacy. This approach has promising applications in fields like medicine and transportation, offering high-speed encryption and secure cloud storage while maintaining excellent image classification performance.