SUBPIXEL SCANNING FOR SPECTRUM IMAGE ANALYSIS
Subject and Purpose. In this paper, an algorithm is developed to improve the spectrum image resolution in the atomic emission spectrum registration using multi-element detectors and taking multiple exposures with image shifting by a fractional part of a pixel. Methods and Methodology. Mathematical m...
Gespeichert in:
Veröffentlicht in: | Radio physics and radio astronomy (Print) 2024-03, Vol.29 (1), p.38-45 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Subject and Purpose. In this paper, an algorithm is developed to improve the spectrum image resolution in the atomic emission spectrum registration using multi-element detectors and taking multiple exposures with image shifting by a fractional part of a pixel. Methods and Methodology. Mathematical modeling techniques are employed in the coordinate representation dictated by the prior information. The modeling results are checked by experiment. Results. The importance of the resolution value of spectral image registration systems and its impact on the quality of the obtained results have been highlighted. The spectral informativeness of the regularly (matrix- or linearly) structured multi-element detectors has been evaluated against the irregular-structure sensors, such as photographic emulsions. As it has been found that the process of pixel size reduction is demanding, an alternative approach through subpixel image shifting has been proposed. Unlike the case of the corresponding instruments implemented as a rule in the frequency domain, here the relevant theoretical problem reduces to the over- determined system of linear equations. The proposed solution algorithm employs the least squares method having regard to the prior data specificity posed by the isolation of the spectral lines and a smooth flow of the background between them. The digital simulation results and the real experimental evidence from the atomic emission spectral analysis have been presented, illustrating the algorithm operation. The experimental research was carried out for one-dimensional spectrum images registered with linear charge-coupling photodetectors. The so gained resolution was twice the spatial resolution of the photodetector. Conclusions. It has been demonstrated that controlled image shifting not only improves the photometry accuracy but also significantly enhances the detector resolution. The general comparisons of the theory with the experiment have proved the feasibility of bringing the proposed methodology into atomic emission analysis. |
---|---|
ISSN: | 1027-9636 2415-7007 |
DOI: | 10.15407/rpra29.01.038 |