A Spectral Reconstruction Algorithm of Miniature Spectrometer Based on Sparse Optimization and Dictionary Learning.

Area of Science:

• Optical Engineering
• Spectroscopy
• Signal Processing

Background:

• Miniaturizing spectrometers offers significant academic and industrial potential.
• Filter-based miniaturization using broadband filters is a viable approach.
• Spectral reconstruction in filter-based systems can be mathematically modeled as linear equations.

Purpose of the Study:

• To propose a novel spectral reconstruction algorithm for filter-based miniature spectrometers.
• To enhance spectral reconstruction accuracy using sparse optimization and dictionary learning.

Main Methods:

• Developing a spectral reconstruction algorithm based on sparse optimization.
• Incorporating dictionary learning to improve sparsity for non-directly sparse spectra.
• Designing and implementing a prototype filter-based miniature spectrometer for experimental validation.

Main Results:

• Sparse optimization proved effective for spectral reconstruction, regardless of initial spectral sparsity.
• Dictionary learning successfully enhanced the sparsity of non-directly sparse spectra.
• Experimental results validated the feasibility of the proposed reconstruction algorithm.

Conclusions:

• The proposed sparse optimization and dictionary learning approach is well-suited for spectral reconstruction in miniature spectrometers.
• This method shows promising application prospects for the development of practical miniature spectrometers.