A Lightweight Privacy-Preserving System for the Security of Remote Sensing Images on IoT

The acquisition of massive remote sensing data makes it possible to deeply fuse and artificial intelligence (AI). mobility cost advantages new platforms in the Internet Things (IoT) make them ideal for continuous deployment rather than traditional airborne platforms. However, devices are vulnerable malicious attacks privacy leaks when sharing due complex architecture heterogeneity IoT lack a unified security protection mechanism. Traditional methods based on public-key encryption require not only operations but also energy consumption, which poses challenges resources-limited IoT. objective this paper was propose lightweight privacy-preserving system remote-sensing images visual cryptography. This stacking-to-see feature cryptography enables efficient big such as high-resolution multi-scale resource-constrained To alleviate image quality degradation cryptography, we combined denoising neural networks extract high-quality from encrypted datasets, thus improving recognition accuracy loss datasets. We conducted extensive experiments, results verify effectiveness proposed method terms classification accuracy.