Computational resolution enhancement for dispersive spectrometers based on GMC spectral reconstruction

Area of Science:

• Spectroscopy
• Computational Imaging
• Signal Processing

Background:

• Conventional dispersive spectrometers face limitations in spectral resolution due to optical broadening, instrumental response, and detector sampling.
• Achieving high-resolution spectral measurements typically requires expensive and complex hardware modifications.

Purpose of the Study:

• To present a generalized minimax-concave-based spectral reconstruction strategy (GMC-SR) for enhancing the effective spectral resolution of existing grating spectrometers.
• To demonstrate the capability of GMC-SR to improve spectral feature recovery and reduce reconstruction errors without altering spectrometer hardware.

Main Methods:

• Developed a generalized minimax-concave-based spectral reconstruction strategy (GMC-SR).
• Integrated a nonconvex penalty into a deconvolution framework to suppress noise and artifacts while preserving spectral structures.
• Validated the method through simulations and experimental measurements using a compact spectrometer.

Main Results:

• Simulations showed a significant reduction in full width at half maximum (FWHM) from 0.35 nm to 0.095 nm and improved dual-wavelength resolution from 0.35 nm to 0.20 nm.
• Experimental results demonstrated a decrease in single-wavelength FWHM from 0.28 nm to 0.14 nm and improved dual-wavelength resolution from 0.28 nm to 0.189 nm.
• GMC-SR achieved a root-mean-square error (RMSE) of 1.0343e-3 in thin-film measurements and outperformed other deconvolution methods in reconstruction fidelity.

Conclusions:

• GMC-SR offers a general, low-cost computational approach to enhance the effective spectral resolution of dispersive spectrometers.
• The method effectively reconstructs both discrete and broadband spectral features with high fidelity.
• GMC-SR presents a viable alternative to hardware upgrades for achieving high-resolution spectral analysis.