A library-based approach to super resolution metrology and localization without imaging

Obtaining nanoscale spatial information without real-space imaging, but from diffraction patterns, is already a valuable tool in metrology We investigate the potential of metasurfaces for nano-optical sensing and metrology with subwavelength resolution. We aim to exploit complex light scattering from a metasurface, programmable illumination, and retrieval of spatial information of a sample from far-field scattering images. We demonstrate an inversion technique based on singular value decomposition whereby we can retrieve the spatial position of a pointlike light source in a plasmon antenna with lambda/50 resolution just on basis of measured far field radiation patterns. Also, we argue that wavefront-shaping applied to plasmon oligomers enables selective generation of sub-diffractive field patterns that could form an optimal and complete basis for spatially-resolved sensing at the nanoscale. Our endeavours use stochastic optimization to choose wavefronts and plasmonic structures.