Vol. 56, Issue 1, pp. 47-59 (2026)

Abstract

This work introduces a novel iterative reconstruction algorithm that enhances low-resolution optical spectrum. The proposed method effectively recovers spectral features that are otherwise undetectable using low- to moderate-resolution interrogators. An adaptive low-pass filter is integrated into error feedback to effectively suppress noise while selectively preserving high-frequency spectral components during iterative reconstruction process. The algorithm’s performance was validated using fiber grating Fabry–Perot interferometer (FPI) sensor spectrum with varying cavity lengths (4, 19, and 25 mm). Using high-resolution spectrum as references, the reconstructed spectrum demonstrated notable improvements, with reductions in ΔMAE (4 mm: –2.42%, 19 mm: –9.37%, and 25 mm: –32.51%) and ΔMSE (4 mm: –4.23%, 19 mm: –17.96%, and 25 mm: –61.22%) compared to the original low-resolution spectrum. Amongst filter methods, an “equiripple” low-pass filter is effective in enhancing the reconstructed spectrum by suppressing noise, resulting in reduced spectral errors (ΔMAE = –6.49%; ΔMSE = –12.97%). The proposed algorithm offers a practical and computationally efficient solution for recovering lost spectral details, particularly in the high-frequency region with minimal error. This capability is especially valuable for enhancing the performance of low-resolution interrogators, allowing their reconstructed spectrum to exhibit characteristics comparable to those obtained with high-resolution systems.