Quantum-secure authentication and robust retrieval for remote sensing

The quantum-secure CBIR scheme which is presented in this research is a fence against unauthorized users and adversarial attacks on cloud environment remote sensor images. The proposed solution is characterized by Quantum Key Distribution, zero-knowledge proof authentication, QCrypt encryption, adversarial trained deep hashing, and robust watermarking. The model was developed with the help of the MLRSNet dataset, where proposed model recorded a remarkable mean average precision of 94.77% that is 10% improvement from the previous deep-hash results while the watermark-extraction accuracy of over 95% was maintained at 35 dB PSNR. The model has been able provide good result with adversarial, replay, and JPEG compression. Even though the computing engine provides military-grade security and forensic accountability, the current compute overhead is the major reason it has limited use in real-time scenarios.