Computing Metasurface Enabled Quantum Phase Distillation

Quantum image distillation aims to extract the signal image from a mixture of the signal and noise images that are indistinguishable in terms of spectrum and polarization, a process that is unachievable with classical methods. However, in contrast to the amplitude image, phase distillation is challenging via direct spatial or temporal correlation of photon pairs. Incorporating with the polarization entanglement of photon pairs, it is demonstrated here that the phase signal can be quickly distilled by using an integrated computing metasurface to solve the Poisson equation. The proposed technique remains robust even with noise levels two orders higher than the signal, with potential applications in quantum communication and cryptography. Based on the present scheme, it also enables the measurement of photon wave function and the achievement of noninterferometric quantum-enhanced quantitative phase imaging. Our work involving the integrated-metasurface analogue computing paves the way for advancing efficient and rapid quantum information and image processing.