Theoretical study on image encryption based on near-field thermal radiation of many-body systems

Under the escalating challenges in information security, encryption technology has become a core defensive measure, yet single digital encryption still faces interception and decryption risks, driving the development of hybrid encryption combining physical mechanisms with digital algorithms. Near-field thermal radiation exhibits high sensitivity to parameters such as materials and distance at the sub-wavelength scales, which provides the possibility of constructing high-security, electromagnetic-interference-resistant physical encryption systems. In this Letter, a near-field thermal radiation modulator is presented based on a vanadium dioxide/graphene emitter and graphene/silicon carbide receiver photonic heterostructure, thereby filling the research gap in the transient analysis of many-body near-field thermal radiation. A key space is established by tuning the emitter temperature, gap distance, and graphene Fermi level. At the algorithmic level, image scrambling and diffusion are achieved by combining a convolutional neural network with chaotic systems. Analysis demonstrates that the proposed encryption system exhibits excellent performance in pixel distribution, correlation, information entropy, and resistance to differential attacks, providing a novel technical pathway for high-security image encryption.