Reconfigurable Snapshot Hyperspectral Imaging Sensor Based on Monochromatic Pattern Match of Gradient Geometry Metasurface

Metasurface-based hyperspectral imaging has profoundly impacted the field, with its easy integration and miniaturization advantages. Notably, metallic metasurfaces have been a research subject in sensing and imaging due to their strong capability of restraining light and supporting high-fidelity and cost-effective manufacturing processes. In this work, a novel metallic metasurface with gradient geometry was proposed for hyperspectral imaging, which highlights the advantages of defect tolerance and reconfigurability of imaging units with a spectral image resolution of 135 × 75 superpixels. An orthogonal matching algorithm with postcalibration based on monochromatic transmittance images of the metasurface was designed for spectral reconstruction. The proposed reconstruction algorithm simplifies calibration and can compensate for systematic losses. The operating bandwidth of the hyperspectral imaging sensor can be reconfigured, and the miniaturized nature makes it potentially integrated into smartphones and other instruments with lightweight computing and fast reconstruction. Moreover, the strategy has been validated in sensing and real-object surface imaging with a spectral resolution of approximately 1.99 nm under a 2 nm line width. Its efficacy is verified in the 400–800 nm band with a high fidelity of up to 86.7%.