Sub-pixel algorithms on linear-array detector grating spectrometer

Abstract

The spectrum of Gauss apodized fiber Bragg grating (FBG) is analyzed by using transfer matrix method, and the basic principles of sub-pixel curve fitting algorithms including mass center, Gauss and full-width-at-half-maximum (FWHM) are discussed. Based on this, simulation and experimental results are given to compare the three curve fitting algorithms. The results reveal that in simulation, mass center fitting method with an accuracy of 13pm while in experiment, FWHM fitting method with an accuracy of 23pm provides a better match to the actual curve. The research achievements have been applied on the prototype of linear-array detector grating spectrometer and have potential market in portable FBG interrogation field.

Hi-Accuracy Method for Spectrum Shift Determination

“…To improve the determination accuracy of the resonance position in the reflection spectrum of fiber Bragg gratings, various approaches have been proposed. These approaches are associated with the methods of sub-pixel processing of spectrum data [16][17][18], which make it possible to increase the resolution by one or one and a half orders of magnitude compared with the physical resolution of a spectrum analyzer [19]. It was reported that the accuracy of the FBG central wavelength determination can be increased by 31 times compared to the spectrometer resolution of 405 pm using the mass center method [17].…”

“…These approaches are associated with the methods of sub-pixel processing of spectrum data [16][17][18], which make it possible to increase the resolution by one or one and a half orders of magnitude compared with the physical resolution of a spectrum analyzer [19]. It was reported that the accuracy of the FBG central wavelength determination can be increased by 31 times compared to the spectrometer resolution of 405 pm using the mass center method [17]. It is also shown that the 3-dB Bandwidth algorithm improves the accuracy by 5.3 times, and the Gauss curve fitting algorithm provides a twofold increase in accuracy.…”