Ultra-wide-angle multispectral narrow-band absorber for infrared spectral reconstruction

Summary This paper presents the design of an ultra-wide-angle multispectral narrow-band absorber for reconstructing infrared spectra. The absorber offers several advantages, including polarization sensitivity, robustness against structural wear, wide azimuthal angle coverage, high narrow-band absorption, and adjustable working wavelength. To accomplish infrared spectrum reconstruction, an absorber is employed as a spectral sampling channel, eliminating the influence of slits or complex optical splitting elements in spectral imaging technology. Additionally, we propose using a truncation regularization algorithm based on the design matrix singular value ratio, namely IReg, which can enable high-precision spectral reconstruction under largely disturbed environments. The results demonstrate that, even when the number of absorption spectrum curve is reduced to a range of 1/2 to 1/3, high-precision spectral reconstruction is achievable for both flat and high-energy steep mid- and long-infrared spectral targets, while effectively accomplishing data dimension reduction.


Supplemental ITEMS
In the manuscript,  presents the analysed relationship between the narrow-band absorption efficiency and duty ratio for structural gradient angles of β=62-119°; detailed data are presented in Table S1.In the manuscript,  and Table S2 showcase the relationship between narrow-band absorption efficiency and period for structural gradient angles of β=62-119°.
Table S1.[ S4-S11.These results indicate that at SNR = 30 dB, significant jitter is observed in the reconstruction curves obtained using the L1 and L2 algorithms; specifically, for the smooth target spectral curves, the reconstruction accuracy is poor.However, the reconstruction performance and accuracy of the IReg algorithm is significantly higher than those of the L1 and L2 algorithms.At an SNR of 20 dB, the IReg algorithm can still reconstruct the six target spectral curves with a high precision, which reflects the anti-noise characteristic of this algorithm.
Data analysis of absorption efficiency and duty cycle at structural gradient angles of β=62-119°], Related to Figure5.TableS1.Data analysis of absorption efficiency and duty cycle at structural gradient angles of β=62[Data analysis of absorption efficiency and period at structural gradient angles of β=62-

Figure S1 .
Figure S1.Comparative analysis of spectral reconstruction results D-85° absorption-type spectral curve data analysis of the accuracy MSE, er, and E of spectral reconstruction of three simulation targets and three ground targets under the wavelength of 6-14 µm interval and 30 dB noise, Related to Figure 10.

Figure S3 .
Figure S3.Comparative analysis of spectral reconstruction results D-100° absorption-type spectral curve data analysis of the accuracy MSE, er, and E of spectral reconstruction of three simulation targets and three ground targets under the wavelength of 6-14 µm interval and 30 dB noise, Related to Figure 10.

Figure S5 .
Figure S5.Comparative analysis of spectral reconstruction results F-105°absorption-type spectral curve data analysis of the accuracy MSE, er, and E of spectral reconstruction of three simulation targets and three ground targets under the wavelength of 6-14 µm interval and 30 dB noise, Related to Figure 10.

Figure S7 .
Figure S7.Comparative analysis of spectral reconstruction results F-100° absorption-type spectral curve data analysis of the accuracy MSE, er, and E of spectral reconstruction of three simulation targets and three ground targets under the wavelength of 5-14 µm interval and 30 dB noise, Related to Figure 10.

Figure S8 .
Figure S8.Comparative analysis of spectral reconstruction results F-100° data analysis of the spectral reconstruction evaluation accuracy MSE, er, E and three ground targets in the 5-14 µm wavelength interval and at 20 dB noise, Related to Figure 11.Table S11.[Absorption spectral curves of the F-100° group and spectral reconstruction data analysis under a 5-14 µm interval and 20 dB noise], Related to Figure 11.

Table S3 .
[Wavelength range and number of spectral curves for different structural gradient angles], Related to Figures5 and 6.

Table S4 .
[Absorption spectral curves of the D-85° group and spectral reconstruction data analysis under a 6-14 µm interval and 30 dB noise], Related to Figure10.

Comparative analysis of spectral reconstruction results
D-85° data analysis of the spectral reconstruction evaluation accuracy MSE, er, E and three ground targets in the 6-14 µm wavelength interval and at 20 dB noise, Related to Figure11.

Table S5
. [Spectral curves of the D-85° group and spectral reconstruction data analysis under the 6-14 µm interval and 20 dB noise], Related to Figure 11.

Table S6 .
[Absorption spectral curves of the D-100° group and spectral reconstruction data analysis under a 6-14 µm interval and 30 dB noise], Related to Figure10.

Comparative analysis of spectral reconstruction results
D-100° data analysis of the spectral reconstruction evaluation accuracy MSE, er, E and three ground targets in the 6-14 µm wavelength interval and at 20 dB noise, Related to Figure11.

Table S7
. [Absorption spectral curves of the D-100° group and spectral reconstruction data analysis under a 6-14 µm interval and 20 dB noise], Related to Figure 11.

Table S8 .
[Absorption spectral curves of the F-105° group and spectral reconstruction data analysis under a 6-14 µm interval and 30 dB noise], Related to Figure10.

Comparative analysis of spectral reconstruction results F
-105° data analysis of the spectral reconstruction evaluation accuracy MSE, er, E and three ground targets in the 6-14 µm wavelength interval and at 20 dB noise, Related to Figure11.

Table S9 .
[Absorption spectral curves of the F-105° group and spectral reconstruction data analysis under a 6-14 µm interval and 20 dB noise], Related to Figure 11.

Table S10 .
[Absorption spectral curves of the F-100° group and spectral reconstruction data analysis under a 5-14 µm interval and 30 dB noise], Related to Figure10.