Reconfigurable Holography Based on Diffractive Deep Neural Networks

In recent years, metasurface has demonstrated unprecedented capabilities in electromagnetic wave modulation and has achieved several results. In this work, we propose a two-layer metasurface structure, in which arbitrarily different holograms can be realized by rotating the first-layer metasurface, which greatly increases the reusability of the metasurface. Numerical simulation shows that our proposed metasurface structure can realize a multi-channel hologram, which has a wide range of applications in the fields of information encryption, information storage, etc.