Abstract
This paper proposes and demonstrates a flexible long-wave infrared snapshot multispectral imaging system consisting of a simple re-imaging system and a pixel-level spectral filter array. A six-band multispectral image in the spectral range of 8-12 µm with full width at half maximum of about 0.7 µm each band is acquired in the experiment. The pixel-level multispectral filter array is placed at the primary imaging plane of the re-imaging system instead of directly encapsulated on the detector chip, which diminishes the complexity of pixel-level chip packaging. Furthermore, the proposed method possesses the merit of flexible functions switching between multispectral imaging and intensity imaging by plugging and unplugging the pixel-level spectral filter array. Our approach could be viable for various practical long-wave infrared detection applications.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Multispectral imaging technology, which can simultaneously obtain two-dimensional spatial information and spectral information of imaging objects, is a widely used information acquisition technology applied in numerous areas, including food safety [1,2], agricultural monitoring [3,4], and medical diagnosis [5,6]. Due to the long-wave infrared (LWIR) spectral information dependance on the intrinsic radiation of the objects, LWIR multispectral imaging systems have the advantages of all-weather detection and high anti-interference capabilities over that working in visible and near-infrared. LWIR multispectral imaging technology thus plays an indispensable role in identifying targets and has been applied in a great many fields, such as remote sensing [7,8], greenhouse gases and polluting gases detection [9], and mineral exploration [10,11].
In the past decades, LWIR multispectral imaging technology has achieved great developments, and a variety of systems have been proposed, such as tunable filter-based multispectral imaging [12–14], filter wheel-based multispectral imaging [15,16]. Although these methods have been proven to be feasible, it is challenging to possess the characteristic of real-time imaging. Therefore, developing a snapshot LWIR multispectral imaging system is of great importance.
Advancements in micro-nano fabrication techniques have provoked revolutionary changes to optical elements. Optical elements composed of micro-nano structures have been proposed to flexibly modulate the optical parameters of incident light on a flat plane, including phase, amplitude, and polarization [17–20]. Many fantastic devices have been demonstrated, including achromatic metalens [21,22], full-colour hologram [23,24], structural colour [25,26], and invisible cloak [27]. A host of micro-nano structures-based filters, such as Fabry-Perot (FP) microcavity [28–32], nanohole [33–37], and subwavelength grating [38–41], have been proposed. Pixel-level spectral filter arrays (PSFAs) have been demonstrated to be a promising way to snapshot and miniaturized multispectral imaging systems [42–44]. Multispectral imaging systems can be obtained by directly integrating such PSFAs with detector chips, which have the advantages of real-time imaging and simple system compositions [45–47]. To the best of our knowledge, such multispectral imaging systems mainly work in visible and near-infrared [42,44,45–47], while the working band of LWIR is rarely reported. In addition, directly integrating PSFAs and detector chips will immobilize the filter arrays onto detectors, which means that the imaging function is fixed. Therefore, to meet the needs of multiple detection methods, developing a flexible imaging system with switchable functions is of great advantage.
In this work, we experimentally demonstrate a flexible LWIR snapshot multispectral imaging system composed of a PSFA and a simple re-imaging system [48]. The proposed imaging system can acquire six-band multispectral images in the spectral range of 8-12 µm with full width at half maximum (FWHM) of about 0.7 µm each band. In our system, the PSFA is placed at the primary imaging plane of the re-imaging system. In contrast to the previous multispectral imaging methods encapsulating the PSFA with the detector chip, our method diminishes the pixel-level chip packaging process. Furthermore, our system provides flexible functions switching between multispectral imaging and intensity imaging by plugging and unplugging the PSFA.
2. Results and discussion
2.1 Design and simulations
The proposed LWIR snapshot multispectral imaging system, consisting of a simple re-imaging system and a PSFA, is shown in Fig. 1. The operating principle of the imaging system is that the spectral information of the object is first filtered out by the PSFA placed at the primary image plane, and then imaged on the detector by the relay lens. The image of each spectral band can be extracted from the multispectral image received by the detector. In addition, the PSFA in the system can be removed flexibly, in which case the system possesses the function of intensity imaging.
The key element of the system is PSFA. As depicted in Fig. 2(a), the supercell of the PSFA consists of six sub-unit filter cells arranged by 2 × 3. Here, the filter of each spectral band consists of seven layers of dielectric materials, alternating high refractive index material layers and low refractive index material layers, forming FP microcavity on the substrate. Considering the operating band, the BaF2 is selected as the substrate. Ge is selected as the high refractive index material with a refractive index of 4.0 in the operating band, and ZnS as the low refractive index material with a refractive index of 2.2 [49]. The combination of Ge and ZnS is symmetrically distributed on both sides of the ZnS spacer layer as the Bragg reflectors. Six spectral filters with different spacer layer thicknesses are selected as the sub-unit cells of the PSFA. The thickness of different layers (ti, i = 1, 2, …7) for each spectral band is shown in Table 1.
The transmission spectra of six bands were calculated using the commercial software CST Microwave Studio, as plotted in Fig. 2(b). The spectral peaks range from 8 to 12 µm and show transmittance around 90% in all spectral bands. Accurately, the central wavelength of each band is designed to be 8.7 µm, 9.3 µm, 10 µm, 10.8 µm, 11.3 µm and 11.9 µm, respectively. Moreover, the FWHM of each band is around 0.7 µm.
2.2 Fabrication and characterization
In order to verify the design, three types of PSFAs with pixel sizes of w = 25 µm, w = 35 µm and w = 50 µm are fabricated by overlay, with the corresponding pixel number of 324 × 256, 324 × 256 and 162 × 128, respectively. The fabrication processes are illustrated in Fig. 3(a).
First, the bottom Bragg reflector and the spacer layer are evaporative deposited on the BaF2 substrate with diameter of 25.4 mm. Then followed by three steps of the overlay process. Each step of the overlay process consists of four procedures: spin coating 0.5 µm AZ1500 photoresist, exposure by direct laser writing (DLW) and development, ion beam etching (IBE), and photoresist removing. In the first step of the overlay process, the areas to be etched consists of two parts, the regions of PSFA with red dotted borders and the regions of cross-alignment marks with blue dotted borders. The cross-alignment marks are located around the PSFA with fixed distance to the boundary of PSFA and size of 20 µm. In each subsequent step of the overlay process, the alignment of the exposure area should be completed according to the alignment marks. The regions of PSFA with red dotted borders are the areas to be etched in the following two steps of the overlay process. The etch thickness of the three overlay steps is 0.2 µm, 0.4 µm and 0.4 µm, respectively. Finally, the top Bragg reflector layers are obtained by evaporative deposition, and the PSFA consisting of six bands is obtained.
As depicted in Fig. 3(b), the three-dimensional contour image of the processed device is measured using a laser scanning confocal microscope (LSCM) to verify the accuracy of our fabrication processes. It can be seen that the relative height of the six filters is 0.26 µm, 0.44 µm, 0.65 µm, 0.83 µm, 1.04 µm and 1.22 µm, respectively. Considering the fabrication error, the height difference between each filter is basically the same as the design.
The characteristic spectral response of the fabricated devices is measured using the Fourier transform infrared (FTIR) microscope (HYPERION, Bruker). The PSFAs are placed on the object stage of the microscope. The spot size of the microscope is fixed and larger than the pixel size of filter array. Due to the small measurement area, the incident light is almost vertical. What really affects the result is the effective measurement area. To ensure that only one pixel is measured at a time, the effective measurement area should be adjusted slightly smaller than the pixel size when measuring different PSFAs. Figure 4 shows the transmission spectrum of each band for PSFAs with different pixel sizes. For the PSFA with a pixel size of 35 µm, the central wavelength of each band is 8.9 µm, 9.7 µm, 10.3 µm, 10.9 µm, 11.5 µm and 11.9 µm, respectively. The difference in the central wavelength between the experiment and the simulation may be caused by the fabrication error. The FHWM of each spectral band of the fabricated PSFAs is about 0.7 µm, agreeing well with the simulation results. It is worth noting that the transmission efficiency is increased with w. The transmission efficiency for the PSFA with a pixel size of 25 µm is about 40%, while for PSFAs with a pixel size of 35 µm and 50 µm, the transmission efficiency is about 60% and 80%, respectively. The transmission efficiency is increased by increasing w, which the diffraction principle can explain. Diffraction occurs when the incident light passes through the PSFAs. Owing to the w increases, the diffraction effect weakened.
2.3 Multispectral imaging experiment
To demonstrate multispectral imaging, a LWIR uncooled detector (HiNet VOx microbolometer) with a pixel size of 25 µm and a pixel number of 324 × 256 was used in the experiment. Comprehensively considering the transmittance and pixel number, the PSFA whose pixel size is 35 µm and pixel number is 324 × 256 was finally selected for the imaging experiment.
Prior to the multispectral imaging experiment, pixel alignment between the PSFA and detector is carried out first. The setup is schematically shown in Fig. 5(a). A blackbody radiation source was used as the imaging object. A spectral filter with the same central wavelength as band 1 of the PSFA was used to realize pixel alignment. The relay lens is an objective with diameter of 25.4 mm, length of 18.8 mm and focal length of 13 mm.
The PSFA and relay lens were placed on the XYZ-translation stage, and pixel alignment was achieved by adjusting the location of the PSFA and the relay lens. To ensure the corresponding relationship between PSFA and pixels of detector, the transversal magnification of PSFS through relay lens is 0.7143 (equal to the ratio of the pixel size of the detector and PSFA, that is 25/35). The pixel alignment result is depicted in Fig. 5(b). When the pixel alignment is completed, each pixel of the detector corresponds to only one sub-unit cell of the PSFA. It can be seen from the partially enlarged drawing that the intensity of spectral band 1 is larger than other bands because of the existence of the filter, indicating that the pixel alignment is realized.
Following the pixel alignment, the multispectral imaging experiment was carried out. A fore lens was placed at the front of PSFA at a certain distance to form a simple re-imaging system. The fore-lens is a ZnSe flat-convex lens with diameter of 25.4 mm and focal length of 25.4 mm. The setup is schematically shown in Fig. 6(a). The physical picture of the imaging system is illustrated by Fig. 13 in Appendix C. A multispectral imaging target containing six letters is designed for the imaging experiment. The six letters are formed by filters that work in different spectral bands. The central working wavelengths of letters “I”, “O”, “E”, “C”, “A”, and “S” are chosen as 8.9 µm, 9.7 µm, 10.3 µm, 10.9 µm, 11.5 µm and 11.9 µm, respectively, which corresponds to the operating wavelength of six sub-unit cells of PSFA. The recorded multispectral image with a pixel number of 324 × 256 is illustrated in Fig. 6(b). The red border area of the partially enlarged drawing represents a 2 × 3 pixel array. For each letter in the multispectral imaging, only the pixels corresponding to the PSFA are lit, which means the designed PSFA works. For the letter “I”, only the pixels corresponding to spectral band 1 have larger intensity. Meanwhile, for letters “O”, “E”, “C”, “A”, and “S”, the pixels with larger intensity correspond to spectral bands 2, 3, 4, 5, and 6, respectively.
Furthermore, the spectral information of the imaging target can be extracted from the recorded multispectral image. It is worth noting that there is information loss when directly extracting spectral information, owning to the pixel-based spectral acquisition. In this work, the pixel number of the directly extracted image at each band is about 108 × 128, as shown in Fig. 7(a). Aiming at the problem of pixel missing caused by plugging the PSFA, the demosaicking algorithm [50,51] was used to reconstruct the missing pixels by interpolation. The images of six spectral bands have a pixel number of 324 × 256 are finally obtained, as shown in Fig. 7(b). To further illustrate the necessity of missing pixel reconstruction, the image of the letter “A” before and after reconstruction is enlarged, as shown in Fig. 7(c). It can be seen that due to the higher pixel number, the image after missing pixel reconstruction is clearer.
Owing to a re-imaging system, our system possesses the advantage of switching the imaging functions between multispectral imaging and intensity imaging by plugging and unplugging the PSFA. The intensity imaging result after removing the PSFA is plotted in Fig. 8(a). On account of the imaging system possessing higher light efficiency when the PSFA is unplugged, the intensity image has higher contrast than the multispectral image, which has an advantage in resolving the details of the imaging targets. Compared with the multispectral imaging, the enlarged letter “A” of the intensity image has a higher image contrast and sharper image detail, as shown in Fig. 8(b).
3. Conclusion
In conclusion, a flexible LWIR snapshot multispectral imaging system is proposed and experimentally demonstrated. The multispectral imaging system is achieved using a simple re-imaging system and a PSFA. As a proof-of-concept, an imaging system acquiring six-band multispectral images in the spectral range from 8-12 µm is finally verified. The FWHM of each band is about 0.7 µm. The experimental results indicate that the six letters “I”, “O”, “E”, “C”, “A”, and “S” with different center wavelengths can be accurately imaged on the corresponding spectral band. Owing to the use of the re-imaging system, the PSFA is placed at the primary imaging plane of the re-imaging system rather than directly integrating with the detector chip, which diminishes the complexity of pixel-level chip packaging. Furthermore, this approach possesses the merits of switching the imaging functions between multispectral and intensity imaging, which is suitable for diversified LWIR detection applications. In addition, the proposed flexible imaging system can potentially switch other imaging functions by only change the PSFA, such as polarization imaging [52].
Appendix A: optical and mechanical design
The relay lens used in the experiment is an objective. The distortion of different wavelengths and fields of view are illustrated in Fig. 9. For wavelengths of 8 µm, 10 µm, and 12 µm, the maximum distortion is -1.68% in the field of view range of 12.5°, indicating that the relay lens has a fine imaging quality.
A filter wheel is used to switch the imaging functions between multispectral and intensity imaging, which different from the filter wheel used to realize spectral imaging [15], as shown in Fig. 10. The filter wheel owns two through holes, one to hold the PSFA for spectral imaging and the other to hold BaF2 substrate for intensity imaging. Two grooves on the back of the filter wheel for initial alignment when switching imaging functions.
Appendix B: the transmission spectra of the filter and each letter
In the imaging experiment, a spectral filter was used to realize pixel alignment. The filter is based on FP cavity, with the same layer number, material, and thickness as band 1 of the PSFA. The transmission spectrum of the filter is illustrated in Fig. 11. The center wavelength of the filter is 8.9 µm, as same as band 1.
To verifiably illustrate multispectral imaging capabilities, a target containing six letters is designed and fabricated for the imaging experiment. The letters consist of subwavelength Au holes with different period P and diameter D on BaF2 substrate, as illustrated in Fig. 12(a) and Table 2. Each letter has a dedicated working band. The central working wavelengths of letters “I”, “O”, “E”, “C”, “A”, and “S” are chosen as 8.9 µm, 9.7 µm, 10.3 µm, 10.9 µm, 11.5 µm and 11.9 µm, respectively, which corresponds to the operating wavelength of six sub-unit cells of PSFA. The transmission spectra of six letters are depicted in Fig. 12(b).
Appendix C: photograph of the optical system used for imaging experiment
Funding
National Key Research and Development Program of China (2018YFA0701800, 2021YFA1401000); Sichuan Province Science and Technology Support Program (2019JDJQ0011).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. J. Qin, K. Chao, M. S. Kim, R. Lu, and T. F. Burks, “Hyperspectral and multispectral imaging for evaluating food safety and quality,” J. Food Eng. 118(2), 157–171 (2013). [CrossRef]
2. H. Huang, L. Liu, and M. O. Ngadi, “Recent developments in hyperspectral imaging for assessment of food quality and safety,” Sensors 14(4), 7248–7276 (2014). [CrossRef]
3. A. Laamrani, A. A. Berg, P. Voroney, H. Feilhauer, L. Blackburn, M. March, P. D. Dao, Y. He, and R. C. Martin, “Ensemble identification of spectral bands related to soil organic carbon levels over an agricultural field in Southern Ontario, Canada,” Remote Sens. 11(11), 1298 (2019). [CrossRef]
4. C. Wang, B. Liu, L. Liu, Y. Zhu, J. Hou, P. Liu, and X. Li, “A review of deep learning used in the hyperspectral image analysis for agriculture,” Artif. Intell. Rev. 54(7), 5205–5253 (2021). [CrossRef]
5. G. Lu and B. Fei, “Medical hyperspectral imaging: a review,” J. Biomed. Opt. 19(1), 010901 (2014). [CrossRef]
6. W. R. Johnson, D. W. Wilson, W. Fink, M. Humayun, and G. Bearman, “Snapshot hyperspectral imaging in ophthalmology,” J. Biomed. Opt. 12(1), 014036 (2007). [CrossRef]
7. A. N. French, T. J. Schmugge, J. C. Ritchie, A. Hsu, F. Jacob, and K. Ogawa, “Detecting land cover change at the Jornada Experimental Range, New Mexico with ASTER emissivities,” Remote Sens. Environ. 112(4), 1730–1748 (2008). [CrossRef]
8. M. L. McDowell and F. A. Kruse, “Enhanced compositional mapping through integrated full-range spectral analysis,” Remote Sens. 8(9), 757 (2016). [CrossRef]
9. M. Gålfalk, G. Olofsson, P. Crill, and D. Bastviken, “Making methane visible,” Nat. Clim. Chang. 6(4), 426–430 (2016). [CrossRef]
10. Y. Ninomiya and B. Fu, “Thermal infrared multispectral remote sensing of lithology and mineralogy based on spectral properties of materials,” Ore Geol. Rev. 108, 54–72 (2019). [CrossRef]
11. G. Notesco, V. Kopačková, P. Rojík, G. Schwartz, I. Livne, and E. B. Dor, “Mineral classification of land surface using multispectral LWIR and hyperspectral SWIR remote-sensing data. A case study over the Sokolov Lignite Open-Pit Mines, the Czech Republic,” Remote Sens. 6(8), 7005–7025 (2014). [CrossRef]
12. H. Mao, K. K. M. B. D. Silva, M. Martyniuk, J. Antoszewski, J. Bumgarner, B. D. Nener, J. M. Dell, and L. Faraone, “MEMS-based tunable Fabry–Perot filters for adaptive multispectral thermal imaging,” J. Microelectromech. S. 25(1), 227–235 (2016). [CrossRef]
13. N. Gupta, “Materials for imaging acousto-optic tunable filters,” Proc. SPIE 9100, 91000C (2014).
14. H. Zhang, A. Muhammad, J. Luo, Q. Tong, Y. Lei, X. Zhang, H. Sang, and C. Xie, “MWIR/LWIR filter based on Liquid–Crystal Fabry–Perot structure for tunable spectral imaging detection,” Infrared Phys. Technol. 69, 68–73 (2015). [CrossRef]
15. D. Cabib, M. Lavi, A. Gil, E. Ohel, J. Dolev, and U. Milman, “A long wave infrared (LWIR) spectral imager (7.7 to 12.3 µ) based on cooled detector array and high resolution circular variable filter (CVF),” Proc. SPIE 8896, 88960R (2013).
16. J. Brauers, N. Schulte, and T. Aach, “Multispectral filter-wheel cameras: geometric distortion model and compensation algorithms,” IEEE Trans. on Image Process. 17(12), 2368–2380 (2008). [CrossRef]
17. N. Yu and F. Capasso, “Flat optics with designer metasurfaces,” Nat. Mater. 13(2), 139–150 (2014). [CrossRef]
18. X. Luo, “Subwavelength artificial structures: opening a new era for engineering optics,” Adv. Mater. 31(4), 1804680 (2019). [CrossRef]
19. X. Ma, M. Pu, X. Li, Y. Guo, and X. Luo, “All-metallic wide-angle metasurfaces for multifunctional polarization manipulation,” Opto-Electron. Adv. 2(3), 18002301 (2019). [CrossRef]
20. Z. Yue, J. Li, J. Li, C. Zheng, J. Liu, G. Wang, H. Xu, M. Chen, Y. Zhang, Y. Zhang, and J. Yao, “Terahertz metasurface zone plates with arbitrary polarizations to a fixed polarization conversion,” Opto-Electron. Sci. 1(3), 210014 (2022). [CrossRef]
21. S. Wang, P. C. Wu, V.-C. Su, Y.-C. Lai, M.-K. Chen, H. Y . Kuo, B. H. Chen, Y. H. Chen, T.-T. Huang, J.-H. Wang, R.-M. Lin, C.-H. Kuan, T. Li, Z. Wang, S. Zhu, and D. P. Tsai, “A broadband achromatic metalens in the visible,” Nat. Nanotechnol. 13(3), 227–232 (2018). [CrossRef]
22. Y. Wang, Q. Fan, and T. Xu, “Design of high efficiency achromatic metalens with large operation bandwidth using bilayer architecture,” Opto-Electron. Adv. 4(1), 200008 (2021). [CrossRef]
23. F. Zhang, M. Pu, P. Gao, J. Jin, X. Li, Y. Guo, X. Ma, J. Luo, H. Yu, and X. Luo, “Simultaneous full-color printing and holography enabled by centimeter-scale plasmonic metasurfaces,” Adv. Sci. 7(10), 1903156 (2020). [CrossRef]
24. X. Li, L. Chen, Y. Li, X. Zhang, M. Pu, Z. Zhao, X. Ma, Y. Wang, M. Hong, and X. Luo, “Multicolor 3D meta-holography by broadband plasmonic modulation,” Sci. Adv. 2(11), e1601102 (2016). [CrossRef]
25. W. Yang, S. Xiao, Q. Song, Y. Liu, Y. Wu, S. Wang, J. Yu, J. Han, and D.-P. Tsai, “All-dielectric metasurface for high-performance structural color,” Nat. Commun. 11(1), 1864 (2020). [CrossRef]
26. R. Fu, K. Chen, Z. Li, S. Yu, and G. Zhang, “Metasurface-based nanoprinting: principle, design and advances,” Opto-Electron. Sci. 1(10), 220011 (2022). [CrossRef]
27. Y. Yang, L. Jing, B. Zheng, R. Hao, W. Yin, E. Li, C. M. Soukoulis, and H. Chen, “Full-polarization 3D metasurface cloak with preserved amplitude and phase,” Adv. Mater. 28(32), 6866–6871 (2016). [CrossRef]
28. A. McClung, S. Samudrala, M. Torfeh, M. Mansouree, and A. Arbabi, “Snapshot spectral imaging with parallel metasystems,” Sci. Adv. 6(38), eabc7646 (2020). [CrossRef]
29. L. Pjotr Stoevelaar, J. Berzins, F. Silvestri, S. Fasold, K. Z. Kamaki, H. Knopf, F. Eilenberger, F. Setzpfandt, T. Pertsch, S. M. B. Baumer, and G. Gerini, “Nanostructure-modulated planar high spectral resolution spectro-polarimeter,” Opt. Express 28(14), 19818–19836 (2020). [CrossRef]
30. A. M. Shaltout, J. Kim, A. Boltasseva, V. M. Shalaev, and A. V. Kildishev, “Ultrathin and multicolour optical cavities with embedded metasurfaces,” Nat. Commun. 9(1), 2673 (2018). [CrossRef]
31. J. Zhao, M. Qiu, X. Yu, X. Yang, W. Jin, D. Lei, and Y. Yu, “Defining deep-subwavelength-resolution, wide-color-gamut, and large-viewing-angle flexible subtractive colors with an ultrathin asymmetric Fabry–Perot lossy cavity,” Adv. Opt. Mater. 7(23), 1900646 (2019). [CrossRef]
32. Z. Yang, Y. Chen, Y. Zhou, Y. Wang, P. Dai, X. Zhu, and H. Duan, “Microscopic interference full-color printing using grayscale-patterned Fabry–Perot resonance cavities,” Adv. Opt. Mater. 5(10), 1700029 (2017). [CrossRef]
33. Y. Shah, P. Connolly, J. Grant, D. Hao, C. Accarino, X. Ren, M. Kenney, V. F. Annese, K. G. Rew, Z. Greener, Y. Altmann, D. Faccio, G. S. Buller, and D. R. S. Cumming, “Ultralow-light-level color image reconstruction using high-efficiency plasmonic metasurface mosaic filters,” Optica 7(6), 632–639 (2020). [CrossRef]
34. X. Wu, D. Gao, Q. Chen, and J. Chen, “Multispectral imaging via nanostructured random broadband filtering,” Opt. Express 28(4), 4859–4875 (2020). [CrossRef]
35. A. Wang and Y. Dan, “Mid-infrared plasmonic multispectral filters,” Sci. Rep. 8(1), 11257 (2018). [CrossRef]
36. B. J. Craig, J. Meng, V. Shrestha, J. J. Cadusch, and K. B. Crozier, “Mid- to long-wave infrared computational spectroscopy using a subwavelength coaxial aperture array,” Sci. Rep. 9(1), 13537 (2019). [CrossRef]
37. J. Zhao, X. Yu, X. Yang, Q. Xiang, H. Duan, and Y. Yu, “Polarization independent subtractive color printing based on ultrathin hexagonal nanodisk-nanohole hybrid structure arrays,” Opt. Express 25(19), 23137 (2017). [CrossRef]
38. T. Xu, Y. Wu, X. Luo, and L. J. Guo, “Plasmonic nanoresonators for high-resolution colour filtering and spectral imaging,” Nat. Commun. 1(1), 59 (2010). [CrossRef]
39. A. F. Kaplan, T. Xu, and L. J. Guo, “High efficiency resonance-based spectrum filters with tunable transmission bandwidth fabricated using nanoimprint lithography,” Appl. Phys. Lett. 99(14), 143111 (2011). [CrossRef]
40. E. Li, X. Chong, F. Ren, and A. X. Wang, “Broadband on-chip near-infrared spectroscopy based on a plasmonic grating filter array,” Opt. Lett. 41(9), 1913–1916 (2016). [CrossRef]
41. K. Lee, J. Jang, S. J. Park, C. Ji, S. Yang, L. J. Guo, and H. J. Park, “Nanoimprint lithography: angle-insensitive and CMOS-compatible subwavelength color printing,” Adv. Opt. Mater. 4(11), 1696–1702 (2016). [CrossRef]
42. C. Williams, G. S. D. Gordon, T. D. Wilkinson, and S. E. Bohndiek, “Grayscale-to-color: scalable fabrication of custom multispectral filter arrays,” ACS Photonics 6(12), 3132–3141 (2019). [CrossRef]
43. X. Li, Z. J. Tan, and N. X. Fang, “Grayscale stencil lithography for patterning multispectral color filters,” Optica 7(9), 1154–1161 (2020). [CrossRef]
44. J. Lee, Y. Park, H. Kim, Y.-Z. Yoon, W. Ko, K. Bae, J.-Y. Lee, H. Choo, and Y.-G. Roh, “Compact meta-spectral image sensor for mobile applications,” Nanophotonics 11(11), 2563–2569 (2022). [CrossRef]
45. S. P. Burgos, S. Yokogawa, and H. A. Atwater, “Color imaging via nearest neighbor hole coupling in plasmonic color filters integrated onto a complementary metal-oxide semiconductor image sensor,” ACS Nano 7(11), 10038–10047 (2013). [CrossRef]
46. Y. Horie, S. Han, J. Lee, J. Kim, Y. Kim, A. Arbabi, C. Shin, L. Shi, E. Arbabi, S. M. Kamali, H. Lee, S. Hwang, and A. Faraon, “Visible wavelength color filters using dielectric subwavelength gratings for backside-illuminated CMOS image sensor technologies,” Nano Lett. 17(5), 3159–3164 (2017). [CrossRef]
47. Z. Wang, S. Yi, A. Chen, M. Zhou, T. S. Luk, A. James, J. Nogan, W. Ross, G. Joe, A. Shahsaf, K. X. Wang, M. A. Kats, and Z. Yu, “Single-shot on-chip spectral sensors based on photonic crystal slabs,” Nat. Commun. 10(1), 1020 (2019). [CrossRef]
48. Y. Xie, C. Liu, S. Liu, W. Song, and X. Fan, “Snapshot Imaging Spectrometer Based on Pixel-Level Filter Array (PFA),” Sensors 21(7), 2289 (2021). [CrossRef]
49. E. D. Palik, “Handbook of optical constants of solids,” Elsevier, (1997).
50. D. Kiku, Y. Monno, M. Tanaka, and M. Okutomi, “Minimized-Laplacian residual interpolation for color image demosaicking,” Proc. SPIE 9023, 90230 L (2014).
51. B. Park, H. Lee, and S. Lee, “Demosaicking with adaptive reference range selection,” J. Electron. Imag. 28(01), 1 (2019). [CrossRef]
52. T. Xingzhou, O. J. Spires, X. Tian, N. Brock, R. Liang, and S. Pau, “Division of amplitude RGB full-Stokes camera using micro-polarizer arrays,” Opt. Express 25(26), 33160–33175 (2017). [CrossRef]