Preset calibration of a reconstructive spectrometer by post–assembly training with calibration samples

Daichang Dong, Yulong Zhao, Zhijun Sun

Miniature reconstructive spectrometers have attracted significant attention for their broad potential applications. However, performance of the spectrometers is heavily dependent on their preset calibration procedure and spectral reconstruction algorithms, and conventional calibration methods still face limitations. In this work, we propose and demonstrate a post–assembly calibration strategy for a reconstructive spectrometer module via machine learning training with calibration samples to overcome these challenges. For experimental validation, we fabricated a spectrometer module comprising a metallic Fabry–Perot microcavity array chip and calibration samples consisting of multilayered dielectric structures. The module is trained using the transmission spectra of the calibration samples based on the Orthogonal Matching Pursuit (OMP) algorithm. By leveraging the calibration matrix obtained during training, the inverse problem in conventional spectral reconstruction is transformed into a forward calculation problem. Experimental results indicate that the spectrometer calibrated via this post–assembly training achieves performance comparable to, or even superior to, that of conventional methods, particularly when a larger number of calibration samples are utilized.