Enhancing Secure Image Transmission Through Advanced Encryption Techniques

Secure image transmission over the Internet has become a critical issue as digital media become increasingly vulnerable and multimedia technologies progress rapidly. The use of traditional encryption methods to protect multimedia content is often not sufficient, so more sophisticated strategies are required. As part of this paper, an autoencoder-based chaotic logistic map is combined with convolutional neural networks (CNNs) to encrypt images. As a result of optimizing CNN feature extraction, chaotic logistic maps ensure strong encryption while maintaining picture quality and reducing computational costs. In addition to Mean Squared Errors (MSE), entropy, correlation coefficients, and Peak Signal-to-Noise Ratios (PSNRs), the method shows higher performance. In addition to providing increased security, adaptability, and effectiveness, the results prove the method is resilient to many types of attacks. In this study, CNNs and chaotic systems are combined to improve data security, communication, and image transmission.