Open Access
Add to BibSonomy
Abstract
Infrared imaging provides invisible infrared pictures beyond that of normal human vision. However, common infrared imaging systems are incapable of capturing dynamical phenomena with high resolution and a wide field of view (FoV). The proposed modality purposefully selects the incident light with various instantaneous viewing angles and eventually combine them in different temporal sequences to form a larger FoV image, thus allowing one single cooled focal plane arrays to detect an area that appreciably exceeds what is possible with conventional counterparts. The primary principles of the 3.5 $\sim$ 5 $\mu$m imaging system involve contracting and splitting the FoV with the optical wedges and switching the splitted FoV using the micro-lens arrays, which are realized with the secondary imaging structure. In this technique, the entire FoVs are enlarged from 15$^\mathrm{o}$ to 30$^\mathrm{o}$ both in 2-D and 3-D modality when the resolutions keep at 49 lp/mm. Proof-of-concept results demonstrate significantly our mid-infrared imaging modality performs high-quality images in the FoV of -15$^\mathrm{o}$ to 15$^\mathrm{o}$. The present article has brought us a new class of wide FoV imaging configurations that permit the capture of moving objects in large area, which also preserves the high resolution.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Infrared imaging is a widespread and rapidly evolving tool in a variety of different applications such as condition monitoring, nondestructive testing, thermal mapping, night vision and many more [1]. It is possible to easily visualize the invisible effects of temperature change of targets with distinct advantages including less scattering and far transmission capability [2]. Although a sizable percentage of the current infrared imaging modalities aim at improving the spatial resolution, there are numerous applications that demand improvement in field of view (FoV). However, a fundamental challenge existing in conventional system is the inherent trade-off between spatial resolution and FoV [3–6]. Taking the typical infrared imaging system as an example, the practical FoV is normally less than 10$^\mathrm{o}$ because it is very often compromised at the spatial resolution [7,8]. A new imaging scheme with high spatial resolution and wide FoV will open the path to measure and understand the behaviour of large scale and fast process in infrared domain.
Recently, there are several proposals having been directed to improve the performance of the technique in terms of resolution and viewing angle. Pointing mirror is by far the most widely deployed strategy to overcome the mutually contradictory requirements between high spatial resolution and large FoV [9,10]. The performance of the infrared arrangement with pointing mirror is commonly limited by the severe image distortion in different azimuth and pitching angles. In addition, the achievable scan rate is usually too slow, often tens of Hz, incapable of capturing fast dynamic phenomenon. Another type of infrared imaging device that can simultaneously operate in high-resolution and large FoV mode relies on the multi-lenses stitching scheme [11–13]. One major drawback to this method is highly increasing the volume and cost of the system. Micro-lens arrays (MLAs), which can project a focused image on to a parallel, slanted or curved surface, is a very promising technique for the acquisition and display of images of different scenes [14,15].
MLAs, also referred to as microlenticular arrays or lenslet arrays, allow to extract different properties of light by means of precise processing to increase the optical efficiency and control the refringence as well as diffraction [16]. With its peculiar merits of small-size, light weight, high freedom in design and fast scan rate, utilizations for MLAs have been primarily extended to the optical fields of imaging, beam shaping, telecommunications and sensing [17]. Nowadays, MLAs are enabling technologies and have become a necessity in optical systems for the performance improvement or integration. One main attribute of the MLAs is uniformly splitting an entire incident wavefront into a great of tiny parts and thus concentrating on the focal plane respectively, which eventually forms an optical plane with a series of well-distributed focal points. Therefore, the MLAs should be intentionally developed to enable high resolution and large FoV imaging modality on account of controlling the light beam through moving the focal points on the image plane. The new image can be formed by a couple of images in different FoVs and sequences that are operated by the quick movement of MLAs, which indicates usage of MLAs leads to the creation of new imaging scheme with wider FoV.
This paper discusses fundamental and practical consideration of an entirely new 3.5 $\sim$ 5 $\mu$m imaging system with capability of switching four segments of FoV in 3-D model, which mainly consists of four groups of double optical wedges to contract the FoV, two pieces of MLAs to switch the FoV, and one infrared focal arrays (FPAs) with two-dimensional detector pixel matrix. Offering a tunable micro-displacement at high rate, MLAs are expected to be valuable for improving the FoV through switching the instantaneous incident beam ray and then combing them in different temporal sequences. Sec. 2 describes the conceptual theory and design principle of our proposal in detail. In Sec. 3, we report numerical evaluations for the properties of our proposed imaging system that is capable of exactly increasing the FoV by 2 times in one direction. Furthermore, Sec. 4 demonstrates the success of utilizing the MLAs in the 2 × 2 imaging system that performs the practical images with high resolution. Results of captured pictures and discussion on potential improvement are provided in Sec. 5. Finally, we summarize and conclude our work in Sec. 6.
2. Concept and principle
The principle of impactfully expanding the FoV can be described as follows. This is firstly achieved by efficiently contracting the initial FoV, and then equally dividing overall FoV into several segments that is defined as sub-FoV. A bunch of beam ray in each sub-FoV is collimated to the identical incident angle and regularly located on the aperture. The next step is operating the beam ray in various sub-FoVs by means of consecutively switching on or off aperture and converging the passed ray on the detector at different temporal sequences. It also necessitates careful combination of sub-FoV images that will exhibit wide FoV scene. The procedure to implement the switchable imaging system is conceptually depicted in Fig. 1, which holds potentials to greatly enlarge the FoV compared to the conventional imaging counterparts.
Schematic layout of our proposed infrared imaging system is shown in Fig. 2, which is mainly comprised of optical wedges, front and back objective lenses, MLAs, and an infrared detector. In order to collimate the incoming beam ray and largely contract the original FoV, multiple optical wedges are employed and orderly placed on the aperture stop. In other words, when beam ray enters the imaging system and passes through the optical wedges, it will be equally split to several sections that are dedicated to imaging of sub-FoVs. In addition, the front objective lens (FOL) with telecentric structure is designed to focus the incident ray on the first image plane, which ensures the beam ray in one sub-FoV keep the same range of exiting angle [18]. The core component — MLAs, fully filled on the first image plane, serve as the angle filter to select the incident ray with specific angle to form the secondary image in each sub-FoVs. Most importantly, it should be mentioned that MLAs act as the object of back objective lens (BOL) as well as the image of FOL. Finally, the BOL is used to gather switched beam ray on the detector with a given object-image relationship. The images with sub-FoVs, accurately governed by moving the MLAs, will be sequentially observed on the detector, and eventually combine a new image in entire FoV without sacrificing spatial information and resolution.
Fig. 2. Block diagram of the newly-designed infrared imaging modality. OWs, optical wedges; FOL, front objective lens; BOL, back objective lens; MLAs, micro-lens arrays; D, detector.
3. 1 × 2 switchable imaging modality
Operation of the 1 × 2 switchable imaging system working in 3.5 $\sim$ 5 $\mu$m is illustrated in Fig. 3. The aiming FoV of -15$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$ in Y-axis consists of sub-FoV I (-15$^\mathrm{o}$ $\sim$ 0$^\mathrm{o}$) and sub-FoV II (0$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$), and -6$^\mathrm{o}$ $\sim$ 6$^\mathrm{o}$ FoV in X-axis is fixed at design stage. The imaging sensing device — FPAs, with light-sensing pixels of 320 × 256 and each pixel having 30 × 30 $\mu$m size, are located in the focal plane to convert received incident beam to electrical signal. The total size of FPAs is 9.6 × 7.68 mm, and its diagonal length is 12.3 mm. Moreover, the FPAs are also integrated into an evacuated Dewar to minimize thermal currents and achieve high sensitivity [19]. The specific design parameters of this 1 × 2 switchable imaging system are depicted in Table 1. What needs to be elaborated is the imaging system is built in the secondary imaging structure to strictly match the cold shield, which is located in front of the FPAs and protects the FPAs against the unwanted surrounding infrared radiations [20]. Beyond that, manipulating the incident beam with various FoVs also relies on the secondary imaging modality, realized by accurately placing the MLAs on the first imaging plane for focusing and choosing the light beam.
Fig. 3. 1 × 2 switchable FoV infrared imaging modality. OWs, optical wedges; FOL, front objective lens; BOL, back objective lens; MLAs, micro-lens arrays; CS, cold shield; FPAs, focal plane arrays. The blue and green beam ray represent the sub-FoV I and II in this modality, respectively. The total length of the system is 452.8 mm.
Table 1. Characteristics of 1 × 2 switchable imaging modality.
Optical wedges play an active role of deflecting the incoming light beam, which aims at splitting and contracting the original FoV [21]. We use one optical wedge in combination with another wedge to eliminate the chromatic aberration in wide 3.5 $\sim$ 5 $\mu$m waveband [22]. The double optical wedges consist of first optical wedge with vertex angle of 5.559$^\mathrm{o}$ and second optical wedge with vertex angle of 2.508$^\mathrm{o}$, which are made of Si material and Ge material, respectively. Therefore, the sub-FoV I of -15$^\mathrm{o}$ $\sim$ 0$^\mathrm{o}$ and sub-FoV II of 0$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$ are both contracted to -7.5$^\mathrm{o}$ $\sim$ 7.5$^\mathrm{o}$ when a bundle of light ray enter the optical wedges. For instance, the specification for contracting the sub-FoV I is depicted in Fig. 4(a) and reported with mathematical calculation as follows.
Fig. 4. The module of double optical wedges. (a) Geometric model. The blue solid line, red dash-dotted and green dotted line represent the light beam, primary axis and normal vector, respectively. (b) Design diagram.
When the beam ray enters the double optical wedges, the relationships between incident angles and refraction angles are given by
As can been seen in Fig. 5, the conceptual model and design diagram of FOL as the telecentric lens is shown, which enables collecting beam ray from double optical wedges onto the first image plane. The feature of telecentricity, with the benefits of constant FoV and magnification in respect to depth, is achieved by properly placing the exit pupil at infinity distance. Chromatic aberrations of FOL can be substantially reduced by making compound materials including Si, Ge, ZnS and ZnSe, which have different color-dispersing properties. The beam ray, occupying the same positions on the aperture stop, will be converged at the first image plane with different positions and same aperture angles. Owing to the behavior of FOL, the spatial information of aperture stop is successfully converted to information of aperture angle. Consequently, the incident ray of sub-FoV I (-7.5$^\mathrm{o}$ $\sim$ $7.5^o$) and sub-FoV II (-7.5$^\mathrm{o}$ $\sim$ $7.5^o$) is turned into the ray that is focused on the first image plane ranging from -10$^\mathrm{o}$ $\sim$ 0$^\mathrm{o}$ and 0$^\mathrm{o}$ $\sim$ 10$^\mathrm{o}$, respectively.
Fig. 5. The module of FOL with image-space telecentric system. (a) Geometric model. The FoV of FOL 2$\omega _2$ is 15$^\mathrm{o}$, focal length f$_2'$ is 100.87 mm, F-number is 2.83564, first image height 2y$_2'$ is 26.56 mm, and diameter of aperture stop is 34 mm. (b) Design diagram.
As the crucial component in the switchable infrared imaging modality, MLAs are constructed in the Kepler telescope type and comprise a pularity of tiny single lens with a standard 83 × 83 $\mu$m configuration. For the sake of corresponding the pixels on FPAs one by one, the MLAs are manufactured to 320 × 256 units and each unit also has squared shape. In fact, there are two pieces of MLAs used in the system, namely front and back MLAs, respectively, which are shown in Fig. 6. The front MLAs have the capability of managing the incident beam with various angles to different positions on the back MLAs. The back MLAs, consisting of massive units with coated anti-reflection film on the effective aperture and metal reflection film on the complementary area, make the selected beam ray passing through when driven by the piezoelectric ceramics with frequency of 580 Hz. Meanwhile, micro displacement of 26 $\mu$m is governed by back MLAs to switch the incoming beam ray that originates from two sub-FoVs.
Fig. 6. The module of MLAs. (a) Geometric model. The ray from sub-FoV I pass through the MLAs, and ray from sub-FoV II is blocked by the back MLAs. Objective angular aperture is 10$^\mathrm{o}$, and image angular aperture is 5$^\mathrm{o}$. (b) Structure of single unit in back MLAs. (c) Design diagram.
The scenario of BOL to be considered is illustrated in Fig. 7. The next step is imaging the MLAs as a object on the FPAs using the BOL, and the telecentric BOL only accepts incoming beam ray whose principal ray is parallel to the primary axis of the system. The BOL is available with the constant magnification of 0.362 to match the size relationship between MLAs (34 mm) and FPAs (12.3 mm). The F-number of BOL is designed at 2 to satisfy the requirement of cold shield aperture size for optimal performance. Moreover, one cold shield with 10 mm diameter is placed 20 mm behind the BOL to prevent unwanted radiation that will possibly disturb the FPAs. Ultimately, the system is performed as two individual imaging systems with equal FoV, which would be sequentially received by the FPAs (see Fig. 8).
Fig. 7. The module of BOL with object-space telecentric system. (a) Geometric model. The objective angular aperture u$_3$ is 5$^\mathrm{o}$, the image angular aperture u$_3'$ is 14$^\mathrm{o}$, objective height 2y$_3$ is 26.56 mm, and image height 2y$_3'$ is 34 mm. (d) Design diagram.
Fig. 8. The individual infrared imaging system with different sub-FoVs. (a) Sub-FoV I. (b) Sub-FoV II.
As a demonstration of the capability for the switchable FoV imaging modality, we use the modulation transfer function (MTF) and spot diagram to estimate the performance of this proposed system. In Fig. 9 (a) and (b), the cut-off spatial frquency (COSF), defined as the spatial frequency when the MTF drops down to 0.2, is greater than 49 lp/mm, which is closed related with the diffraction limited value. Moreover, the RMS radius of light spot on Fig. 10 (a) and (b) is approximately 6 $\mu$m, which is obviously less than the size of single pixel on FPAs. As shown in Figs. 9 and 10, the achievable MTF curves and spot diagrams prove that two separated sub-FoV systems basically maintain consistent image quality.
Fig. 9. MTFs of 1 × 2 switchable infrared imaging system. (a) Sub-FoV I. COSF=49 lp/mm. (b) Sub-FoV II. COSF=49 lp/mm.
Fig. 10. Spot diagrams of 1 × 2 switchable infrared imaging system. The blue cross, green square, red triangle represent the 3.5 $\mu$m, 4.25 $\mu$m and 5 $\mu$m wavelength of incident light, respectively. The number above the spot stands for the RMS radius in the unit of $\mu$m. (a) Sub-FoV I. (b) Sub-FoV II.
4. 2 × 2 switching image modality
To continue the proof-of-concept that refers to an implementation of enlarging the FoV of infrared imaging system, we explain the approach to switch the FoV in 3-D model utilizing the arrangement shown in Fig. 11. The purposive FoV of -15$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$ in X-axis and -15$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$ in Y-axis is equally separated into four segments, which are defined as sub-FoV I (0$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$ in X-axis and -15$^\mathrm{o}$ $\sim$ 0$^\mathrm{o}$ in Y-axis), sub-FoV II (0$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$ in X-axis and 0$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$ in Y-axis), sub-FoV III (-15$^\mathrm{o}$ $\sim$ 0$^\mathrm{o}$ in X-axis and -15$^\mathrm{o}$ $\sim$ 0$^\mathrm{o}$ in Y-axis) and sub-FoV IV (-15$^\mathrm{o}$ $\sim$ 0$^\mathrm{o}$ in X-axis and 0$^\mathrm{o}$ $\sim$ 15$^\mathrm{o}$ in Y-axis), respectively. The structure of 2 × 2 modality is generally identical with the 1 × 2 model depicted in Fig. 3, expect for the optical wedges and MLAs which are utilized for controlling the beam ray. The FPAs with total size of 8.7 × 8.7 mm, are composed of 290 × 290 units with a pixel pitch of 30 × 30 $\mu$m. Table 2 exhibits the design requirements of the 2 × 2 switchable imaging system in detail.
Fig. 11. 2 × 2 switchable infrared imaging modality. The diameter of entrance pupil is 34 mm, the focal lengths of FOL and BOL are 91.3 mm and 68 mm, respectively.
Table 2. Characteristics of 2 × 2 switchable imaging modality.
Double optical wedges are used effectively to deviate beam ray with a specific angle of 7.5$^\mathrm{o}$, and contract the incident angle scopes both to -7.5$^\mathrm{o}$ to 7.5$^\mathrm{o}$ in four independent sub-FoVs. Figure 12 (a) and (b) depict the cross-sections of optical wedges in Y-Z and X-Z plane, respectively. The detailed information associated with contracting the incident beam angle in four individual sub-FoVs is provided in Fig. 12 (c). Four groups of double optical wedges are located at the upper left, bottom left, right upper and right bottom parts of aperture stop, respectively, which allow to convert the viewing angles to spatial positions in four different sub-FoVs. What is more, it should be strongly emphasized that the angle between principal axis of optical wedges and primary axis of the system is set at 45$^\mathrm{o}$ to ensure implementing identical deflection angle in X-axis and Y-axis.
Fig. 12. Module of optical wedges. (a) Optical wedges in Y-Z plane. (b) Optical wedges in X-Z plane. (c) Model for contracting the sub-FoV I to IV.
24.07 × 24.07 mm MLAs with 290 × 290 units are fabricated in fused silica with each size of 83 × 83 $\mu$m to shape the spatial properties of incident light, which is shown in Fig. 13. After passing the front MLAs, four parts of beam ray with different sub-FoVs are equally distributed in the whole aperture, and simultaneously form four tangent circles with diameter of 26 $\mu$m in X-Y plane. Moving the back MLAs at the step of 26 $\mu$m to permit the ray from one sub-FoV passing through and block the ray from other three sub-FoVs. Figures 14 and 15 compare the imaging performances in four sub-FoVs in terms of MTFs and spot diagrams, indicating that the proposed 2 × 2 switchable infrared imaging modality preserve high resolution and enable capturing dynamic objects in 30$^\mathrm{o}$ FoV.
Fig. 14. MTFs of 2 × 2 switchable infrared imaging system. (a) Sub-FoV I. COSF=49 lp/mm. (b) Sub-FoV II. COSF=49 lp/mm. (c) Sub-FoV III. COSF=49 lp/mm. (d) Sub-FoV IV. COSF=49 lp/mm.
Fig. 15. Spot diagrams of 2 × 2 switchable infrared imaging system. The blue cross, green square, red triangle represent the 3.5 $\mu$m, 4.25 $\mu$m and 5 $\mu$m wavelength of incident light, respectively. The number above the spot stands for the RMS radius in the unit of $\mu$m. (a) Sub-FoV I. (b) Sub-FoV II. (c) Sub-FoV III. (d) Sub-FoV IV.
5. Results and discussion
We verify the performance of the switchable 2 × 2 infrared system with a simulated imaging experiment, which is shown in Fig. 16. The pictures of Fig. 16 (a) $\sim$ (d) illustrate imaging results in the sub-FoV I $\sim$ IV with resolution test chart of USAF-1951, respectively. We can note that switching the FoV is successfully performed with a uniform clarity and resolution in four different sub-FoVs. Hence it can be inferred that the proposed switchable infrared imaging system definitely enlarges the original FoV by 2 times in two direction while keeping high resolution.
Fig. 16. Imaging results of 2 × 2 switchable infrared imaging modality. (a) Sub-FoV I. (b) Sub-FoV II. (c) Sub-FoV III. (d) Sub-FoV IV.
The stray light, from unwanted external infrared radiation, is effectively minimized by the cold shield placed behind the BOL and unable to be detected by the FPAs. Moreover, narcissus effect due to the light reflection from optical and mechanical surfaces is also analyzed in the design procedure [23]. A vast of scattering light comes from surfaces of the BOL and MLAs, and we constrain the cold reflection from those surfaces by increasing the transmission efficiency of related lenses. Our detailed design stage also includes an adequate tolerance analysis on calculating the sensitivity and estimating the performance. Perturbing each element individually and re-optimizing the system are developed to cover all influences on performance degradation [24]. We find that the angle of optical wedges and curvature of FOL are the most sensitive and critical elements that will determine the performance and production cost of this system.
Moving forward, the proposed infrared system is expected to perform satisfactorily in the wide temperature range of -40$^\mathrm{o}$C to 70$^\mathrm{o}$C for diverse practical applications. However, environmental temperature change tends to cause variation of the physical properties of the infrared apparatus, such as refractive index, air space, radii and element thickness, because materials in infrared region have high rates of index variation with temperature. Consequently, temperature change always results in undesirable thermal defocusing and aberrations in the imaging procedure. Here, the MTFs varying in typical operation temperatures are simulated individually and shown in Fig. 17, which obviously indicates the COSF has fallen from 49 lp/mm to 4 lp/mm and its performance can’t meet the actual demanding for applications. One crucial aspect of concern in optical system design is correcting the defocusing and aberration caused by changes in the environmental temperature [25]. Fortunately, simple mechanical athermalization is highly available to the current configuration [26], and moving the FPAs with a tiny displacement ($\Delta$z) in Z-axis is able to completely compensate the severe thermal aberration that is induced by temperature changing. The MTFs almost recover to the initial values by shifting the locations of the FPAs (see Fig. 18).
Fig. 17. MTFs of 2 × 2 switchable infrared imaging system (sub-FoV I) in various working temperatures. (a) -40$^\mathrm{o}$C. COSF=5 lp/mm. (b) -20$^\mathrm{o}$C. COSF=14 lp/mm. (c) 0$^\mathrm{o}$C. COSF=30 lp/mm. (d) 20$^\mathrm{o}$C. COSF=49 lp/mm. (e) 50$^\mathrm{o}$C. COSF=7 lp/mm. (f) 70$^\mathrm{o}$C. COSF=4 lp/mm.
Fig. 18. MTFs of 2 × 2 switchable infrared imaging system (sub-FoV I) with mechanical athermalization. (a) -40$^\mathrm{o}$C. $\Delta$z=-0.270 mm. (b) -20$^\mathrm{o}$C. $\Delta$z=-0.175 mm. (c) 0$^\mathrm{o}$C. $\Delta$z=-0.083 mm. (d) 20$^\mathrm{o}$C. $\Delta$z=0. (e) 50$^\mathrm{o}$C. $\Delta$z=0.150 mm. (f) 70$^\mathrm{o}$C. $\Delta$z=0.242 mm.
6. Conclusion
Infrared imaging is a very indispensable technique for capturing the infrared light from targets and converting it into visible images interpretable by human eyes. It was first developed by the military, and has recently found its way into the commercial market place. Unfortunately, conventional infrared imaging modalities are relatively limited FoV for the sake of keeping certain resolution, and can’t permit clearly observing objects in large area. In this article, we have reported a new experimental technique for enlarging the FoV of infrared imaging system, which brings particular benefits of flexibility, compact and cost-effective by using the MLAs entity. This method features direct switching the FoV horizontally and vertically relying on the common MLAs, overcoming an essential limitation that exists in the typical implementations of imaging system. On the other hand, disadvantages encountering with the MLAs, including process complexity, manufacturing tolerances for the single lenslet and relatively high fabrication cost, are also needed to assign much importance to.
The presented article has covered the principle of 2 × 2 switchable infrared imaging modality, and it also holds considerable potentials to switch N × N imaging system for wider FoV requirement. With its ability to resolve the trade-off between FoV and resolution, the key limiting factor in conventional system, the switchable infrared imaging system is expected to be useful for studying the phenomena in large area and without degrading the image quality. This strategy underscores the need for new concepts leading to revolutionary improvements in integrated image acquisition system by combining multiple sub-FoV images. These accomplishments not only provide valuable insight into the operation of switching FoV, but also serve as a blue print for implementation and optimization of infrared imaging technology.
Funding
Tianjin University (TJU) (2019XZS-0032, 2018XZS-0011).
References
1. M. Vollmer and K. Mollmann, in Infrared thermal imaging, (Wiley-Vch, 2017).
2. R. W. Ruddock, in Basic infrared thermography principles, (Reliabilityweb, 2013).
3. R. C. Hardie, K. J. Barnard, J. G. Bognar, and E. A. Watson, “High-resolution image reconstruction from a sequence of rotated and translated frames and its application to an infrared imaging system,” Opt. Eng. 37(1), 247–260 (1998). [CrossRef]
4. T. Martinez, D. V. Wick, and S. R. Restaino, “Foveated, wide field-of-view imaging system using a liquid crystal spatial light modulator,” Opt. Express 8(10), 555–560 (2001). [CrossRef]
5. G. Korompili, G. Kanakaris, C. Ampatis, and N. Chronis, “A portable, optical scanning microsystem for large field of view, high resolution imaging of biological specimens,” Sens. Actuators, A 279, 367–375 (2018). [CrossRef]
6. N. Kaiser, J. L. Tonry, and G. A. Luppino, “A new strategy for deep wide-field high-resolution optical imaging scanning systems,” Publ. Astron. Soc. Pac. 112(772), 768–800 (2000). [CrossRef]
7. D. Jeong, H. Lee, H. Jeong, C. M. Ok, and H. Park, “Infrared dual-field-of-view optical system design with electro-optic/laser common-aperture optics,” Curr. Opt. Photonics 2, 241–249 (2018).
8. M. B. Vincent and E. V. Ryan, “Simultaneous infrared-visible imager/spectrograph a multi-purpose instrument for the magdalena ridge observatory 2.4 m telescope,” Proc. SPIE 6269, 62692N (2006). [CrossRef]
9. X. Zhang and Y. G. Gao, “Large field of view imaging system for remote target capture and trajectory measurement based on cone rotation,” Rev. Sci. Instrum. 89(6), 063704 (2018). [CrossRef]
10. S. Y. Yi and Y. J. Ko, “Wide field-of-view imaging using a combined hyperbolic mirror,” Curr. Opt. Photonics 1, 336–343 (2017).
11. W. B. Pang and D. J. Brady, “Field of view in monocentric multiscale cameras,” Appl. Opt. 57(24), 6999–7005 (2018). [CrossRef]
12. D. W. Prather, S. Y. Shi, and J. S. Bergey, “Field stitching algorithm for the analysis of electrically large diffractive optical elements,” Opt. Lett. 24(5), 273–275 (1999). [CrossRef]
13. Y. N. Chang, W. Lin, J. Y. Cheng, and S. C. Chen, “Compact high-resolution endomicroscopy based on fiber bundles and imaging stitching,” Opt. Lett. 43(17), 4168–4171 (2018). [CrossRef]
14. H. Yabu and M. Shimomura, “Simple fabrication of micro lens arrays,” Langmuir 21(5), 1709–1711 (2005). [CrossRef]
15. Q. H. Wang, H. Deng, T. T. Jiao, D. H. Li, and F. N. Wang, “Imitating micro-lens array for integral imaging,” Chin. Opt. Lett. 8(5), 512–514 (2010). [CrossRef]
16. C. T. Pan and C. H. Su, “Fabrication of gapless triangular micro-lens array,” Sens. Actuators, A 134(2), 631–640 (2007). [CrossRef]
17. H. W. Lee and B. S. Lin, “Improvement of illumination uniformity for led flat panel light by using micro-secondary lens array,” Opt. Express 20(S6), A788–A798 (2012). [CrossRef]
18. J.-S. Kim and T. Kanade, “Multi-aperture telecentric lens for 3d reconstruction,” Opt. Lett. 36(7), 1050–1052 (2011). [CrossRef]
19. N. Rajic and N. Street, “A performance comparison between cooled and uncooled infrared detectors for thermoelastic stress analysis,” Quant. InfraRed Thermogr. 11(2), 207–221 (2014). [CrossRef]
20. H. B. Xie, Y. P. Su, M. Zhu, L. Yang, S. S. Wang, X. B. Wang, and T. Yang, “Athermalization of infrared optical system through wavefront coding,” Opt. Commun. 441, 106–112 (2019). [CrossRef]
21. C. T. Amirault and C. A. DiMarzio, “Precision pointing using a dual-wedge scanner,” Appl. Opt. 24(9), 1302–1308 (1985). [CrossRef]
22. I. Powell, “Lenses for correcting chromatic aberration of the eye,” Appl. Opt. 20(24), 4152–4155 (1981). [CrossRef]
23. A. S. Lau, “The narcissus effect in infrared optical scanning systems,” Proc. SPIE 107, 590–599 (1976).
24. P. C. Foote and R. A. Woodson, “Lens design and tolerance analysis methods and results,” J. Opt. Soc. Am. 38(7), 590–599 (1948). [CrossRef]
25. D. S. Grey, “Athermalization of optical systems,” J. Opt. Soc. Am. 38(6), 542–546 (1948). [CrossRef]
26. M. Bayar and O. F. Farsakoglu, “Mechanically active athermalization of a forward looking infrared system,” Infrared Phys. Technol. 43(2), 91–99 (2002). [CrossRef]
