Spectral encoding based on narrowband/broadband modulations of QBIC for computational spectral imaging

For computational spectral imaging based on nanophotonic devices, the spectral modulations of the encoders determine the accuracy of the reconstructed spectra and are of crucial importance for various applications. In conventional nanophotonic encoders, broadband transmittance or reflectance modulations are preferred for large wavelength ranges and high light efficiency but meanwhile hamper the spectral resolution and reconstruction fidelity, especially for imaging objects with narrowband spectral features. In this work, a simultaneous narrowband and broadband spectral encoder based on quasi-bound states in the continuum (QBIC) is proposed. By linearly scaling the height-optimal QBIC structure in the lateral dimensions, the electric/magnetic dipole QBIC resonances can cover the entire operating wavelength range. With the combination of broadband transmittance modulation at a high scale factor, the simultaneous narrowband and broadband modulation enhances the average peak signal-to-noise ratio of the reconstructed images by 5.33 dB in real-world scenarios and significantly reduces the error of full width at half maximum by about 5 nm, compared to a typical multilayer film modulator. The concept of spectral encoding with combined broadband and narrowband modulations overcomes the resolution limitations of conventional modulators, offering new possibilities for high-accuracy and miniaturized computational imaging based on nanophotonics.