We report a new dual band compact oblique photography camera (LC11) that is the first to benefit from the incorporation of telecentricity. LC11 has a common front end F/6.6 telescope with 280 mm in aperture that forms its electro-optical (EO, F/7.5) and MWIR (F/5.6) modules. The design allows a substantial reduction in volume and weight due to i) the EO/MWIR compensator and relay lens groups arranged very close to the primary mirror (M1), and ii) light-weighted M1 and SiC main frame (MF) structure. Telecentricity of up to 2 and 0.2 degrees for the EO and MWIR modules, respectively, is achieved by balancing optical power among all lenses. The initial field test shows 0.32 ± 0.05 (EO)/0.20 ± 0.06 (MWIR) in measured MTF at 28 (EO) and 13 (MWIR) cycles/mm in target frequency, and an improved operability with a greater reduction in operational volume and mass than other existing LOROP cameras.
© 2012 OSA
CorrectionsKwang-Woo Park, Jeong-Yeol Han, Jongin Bae, Sug-Whan Kim, Chang-Woo Kim, Hyug-Gyo Rhee, Ho-Soon Yang, and Yun-Woo Lee, "Novel compact dual-band LOROP camera with telecentricity: erratum," Opt. Express 21, 443-443 (2013)
A “Long Range Oblique Photography” (LOROP) camera takes oblique images of distant objects through atmospheric attenuation. Due to significant changes in operational altitude and line of sight (LOS) depression angle, the light intensity arriving at the camera aperture is reduced from high atmospheric absorption . Therefore, cameras of this kind tend to resort to a large aperture front end reflective optical telescope, followed by many refractive and/or reflective relay optical elements for imaging. This makes it necessary for the designers to overcome technical challenges to achieve a light weight and reduction of packaging volume.
We have witnessed the technical evolution of such a camera system over the last several decades, at both visible (0.7–0.9 μm, electro-optical, EO) and infrared (3.7–4.8 μm, MWIR) wavelengths. The LOROP cameras such as CA-295 [2–4], DB-110 [5–9] and a multi-band zoom imager  are fair examples. They employ a common front-end reflective optical system. However, the camera volume becomes relatively large, as the Beam Splitter (BS) is located off the light paths between the primary (M1) and secondary (M2) mirrors, and the aberration compensation lenses are arranged sequentially. A newer design  uses a BS placed between M1 and M2. However, once again, the camera volume becomes large, as the lens compensators are located far from M1 after splitting the visible/near IR and mid-wavelength IR. According to public domain literature, including refereed journal papers, it is not clear where the BS location is in Global Hawk , but it seems that the resulting volume is considerably larger than that of the other examples mentioned previously.
Telecentric optical systems enable uniform magnification and tend to reduce the perspective effects [13,14], even when the object moves forward, backward, left and right, or even when the images are taken from oblique angles. For these reasons, they have been applied to a wide range of optical instrument including metrology [15,16], scanning equipment [17,18], zoom lens systems [19,20], and projection lens systems [21,22] requiring low distortion, and the holographic recording system  taking off axial images with a high signal-to-noise ratio (SNR). For LOROP cameras to take long range oblique images, increased SNR and reduction of scanning distortion and tolerance are essential. However, as of today there has been no study reported in the literature where such telecentricity has been incorporated into the LOROP camera.
In particular, the telecentric optical system has infinitesimal magnification changes over object displacement, thus allowing the relaxation of integration and alignment tolerance to tens of micrometers, as the decenter and despace tolerances can be greatly increased for the image plane. This relaxation in tolerance allows for modular approaches to the complex optical systems in all development phases, including design, manufacturing, integration, and performance test, both in the laboratory and in the field. This modular approach can also reduce instrument maintenance efforts while it is in operation in the field. While we are aware of the fact that such modular design was proposed as a process requirement for the DB-110 camera system development, no specific design process details were available [6,7] and the advantages of modular design have not been presented even in the operational stage after the completion of full development [8,9].
To this extent, this study introduces, for the first time in public archival journal, a new compact and lightweight EO/MWIR LOROP camera incorporating telecentricity. The optical and opto-mechanical design characteristics are described in Chapter 2. Chapter 3 presents technical details of the manufacture, construction and test at the component, module and system levels, including the initial field performance results, before offering some concluding remarks and implications in Chapter 4.
2. Optical and opto-mechanical modeling and simulation
2.1 Optical modeling and simulation
The LC11 system can be divided into three independent modules, which are the front-end optical telescope, EO module including EO compensator lens group, and MWIR module including MWIR compensator and relay lens groups. A diffraction-limited optical design of F/7.5 (EO) and F/5.6 (MWIR) was achieved, and satisfies all the design requirements as listed in Table 1 . The optical design utilizes i) 0.53 x 0.53 (EO) and 0.58 x 0.58 (MWIR) degrees in field of view (FOV), ii) an EO/MWIR common reflective F/6.6 telescope of 280 mm in aperture, iii) the silicon beam splitter (BS) of 116.1 mm in diameter located between M1 and M2, iv) the EO/MWIR imaging lens groups right beside the telescope tube and v) telecentricity at both the EO focal plane and the MWIR intermediate image plane.
Such design characteristics are clearly shown in Fig. 1(a) for EO and (b) for MWIR optical systems. First, the EO module has incident light reflected from M1 and then from M2. Portions of the light are reflected from BS and then by the folding mirror 1 (FM1), before passing through the EO compensator lens group. The light travels to the back scan mirror (BSM) that reflects it to the folding mirror 2 (FM2), before it is focused onto the EO focal plane. Second, the MWIR module has the light refracted through BS and the astigmatism compensator (AC) that corrects the astigmatic wavefront generated from BS . It is then reflected by the folding mirror 3 (FM3), and then the intermediate image plane is formed after passing through the MWIR compensator lens group, two folding mirrors (FM4 and 5), and Back Scan Plate (BSP). The image (already formed) is then re-imaged by folding mirror 6 (FM6) and the relay lens group.
The design utilizes simple conic mirror surfaces for the Cassegrain-type telescope, and uses spherical surfaces for compensator and relay lenses for efficient fabrication. In particular, the EO module has a long back focal length to accommodate BSM and achieve the required telecentricity. The four EO compensator lenses between FM1 and BSM correct residual aberrations from the Cassegrain-type telescope. The MWIR module also requires a lens compensator between FM3 and FM4 for accurate telecentricity, diffraction limited image quality, and long back focal length accommodating BSP at the intermediate image plane. In re-imaging onto the MWIR detector, the purpose of the MWIR relay lens group is to provide the cold stop with a precision control in size and location with respect to the MWIR detector and to move for automatic focus adjustment. It also has a focal length reduction ratio of 0.71, from 2112 mm (the front end telescope and MWIR compensator lens group) to 1500 mm (MWIR relay lens group). Both BSM (for EO) and BSP (for MWIR) compensate the Line of Sight (LOS) movement while scanning the distant target objects. The amplitude of the BSM (θ) is proportional to the FOV of the front-end optics as described in Eq. (1) :
For reduction of the packaging volume, M1 is designed to F/2.9. All EO and MWIR optical elements are arranged to be immediately next to M1, as shown in Fig. 1. The EO/MWIR module including windows and the folding mirrors (FMs) has 11 (EO) and 19 (MWIR) optical components. Together with the obscuration of M2, these result in a transmittance of less than 50%. In general, low transmittance tends to cause SNR degradation, decreasing the overall performance of an optical system (visible from , MWIR from [27,28]). To discourage such SNR degradation, we incorporate telecentricity onto LC11 to increase SNR with the normal incidence of the chief rays . This philosophy leads to additional strengths including less distortion, and eases integration and alignment into a compact space due to relaxed decenter and despace tolerances .
In Cassegrain-type telescope, the telecentricity can be achieved with the compensator optics of positive optical power. The telecentricity condition is defined with the chief ray in parallel to the optical axis in image (detector) plane. Because, for the IR module, the cold stop plays the role of exit pupil and is not at the image plane, the telecentricity at the image plane cannot be realized. This leads to telecentricity achieved in intermediate image plane for both scanning and focusing functions. For the EO module, where the cold stop is not required, we have some degree of freedom to locate the exit pupil at the convenient place. In order to incorporate telecentricity into LC11, we keep the optical power variation among the compensator lenses as small as possible while optimizing the variation in curvature of the radii of adjacent lens surfaces less than 220 mm in the EO/MWIR modules.
Figure 2 shows the compensator lenses of 53.45–340.78 mm (EO) and 21.8–1200 mm (MWIR) in radius of curvature and their locations indicating the optimized distance un-scaled. All lenses are spherical, and the materials used are TAF1, BSC7, and EF2 for the EO module and Si and Ge for the MWIR module. The lenses are smaller than 48/57 mm in diameter for EO/MWIR modules. The resulting telecentricity is about 2 degrees for the EO and 0.2 degrees for the MWIR modules, respectively, leading to an estimated distortion that is smaller than 0.1% (EO) and 0.2% (MWIR).
Aided by the aforementioned telecentricity, the decenter and despace tolerances are both greater than 0.05 mm, which falls within the mechanical tolerance regime. This allows for modular assembly and alignment to the system integration without lengthy optical alignment at the system level. The alignment method for each module and the resulting wave front error (WFE) of the modules are described in Chapter 3.
Figure 3 shows the design modulation transfer function (MTF) of LC11. The on-axis MTF is almost equal to the diffraction-limited performance, which is 0.65 at 28 cycles/mm (EO) and 0.42 at 13 cycles/mm (MWIR). We note that the MTF degradation due to central obscuration of the primary mirror is clear at around 60 (EO) and 15 (MWIR) cycles/mm as shown in Fig. 3.
The stray light analysis was used to find two strong light paths at incident angle of both from −5.6 to −6.6 degrees and from +4.7 to +7.8 degrees among full fields of view as shown in Fig. 4(a) and (b) . The resulting signals to stray light ratios are 0.07% and 1.36% for the two different incident angles without the baffle. According to the simulation results in Fig. 4(c), a baffle incorporated at the back side of M2 suppressed the stray lights to a negligible level. The shape of the baffle is a cone with 73.8 mm in diameter for the wide opening side and 67.8 mm in diameter for the narrow opening side. The detailed dimension of the baffle with yellow line is described in Fig. 4(d).
2.2 Mechanical modeling and simulation
2.2.1 Light-weighted primary mirror and main frame
M1 made of Zerodur has an array of honeycomb-structured holes, with openings toward the rear side only. The holes are 4 different triangular shapes of 41.2, 32.9, 28.8, and 21.6 mm in length, with variable depths ranging from 25 to 30 mm depending on the surface curvature. As a result, M1 is light-weighted to 2.17 kg, achieving about 50% in light-weight ratio. The main frame consists of the baseplate supporting M1, and the tube that holds the M2 cell. We use SiC for the main frame structural material, and the reduced thickness of 10 mm is capable of producing sufficient stiffness. The lightweight SiC mainframe is a cylindrical shape of Φ 340 x 320 mm. Its outer surface is octagonal, whereas the inner surface is circular, to enable high machinability.
2.2.2 Opto-mechanical performance of M1 and M2
The surface deformation of M1 and M2 caused by gravity and thermal gradient environment was computed through finite element analysis using ABAQUSTM. For the input gravity condition, 1 g was given along each of the +X, +Y, and +Z directions. The input thermal condition is the overall temperature change from 20 to 30 °C. The M1 and M2 surface deformations were then analyzed using the Zernike polynomial decomposition . Table 2 summarizes the results, proving that both M1 and M2 satisfy the gravitational and thermal deformation requirements of 20 nm RMS. M1 is supported by three bipod flexures in angular separation of 120°. The flexure is made of Invar, and its dimensions are 25.7 mm in length and 1.6 mm in thickness. Table 3 summarizes the M1 natural frequency and jitter performance prediction.
3. Manufacturing, integration, alignment and testing
3.1 Manufacturing, integration and alignment
M1 was fabricated to 23.5 nm RMS in WFE as shown in Fig. 5(a) . A CGH capable of removing the M1 and M2 residual aberration was used for WFE measurement for the front-end telescope, and the resulting WFE was 44.7 nm RMS as shown in Fig. 5(b). The EO compensator lens group was assembled to the measured WFE of 20.0 nm RMS using another CGH simulating the front-end telescope, as shown in Fig. 5(c). The MWIR relay lens module WFE was measured as 246.0 nm RMS using an interferometer and a reflecting mirror (Fig. 5(d)). The WFE data in Fig. 5 are used as the maintenance qualification criteria for each module.
The EO optical system was then assembled by combining the front-end telescope with the EO module. After assembling the MWIR relay lens group, the MWIR lens compensator group was integrated and aligned while observing the USAF image at the optimal focal point by adjusting the compensation lens (i.e. the first lens in the compensator group after AC).
Figure 6(a) shows the EO system WFE measurement, while the USAF image acquired for the MWIR CTF (Contrast Transfer Function) measurement is presented in Fig. 6(b). The measured WFE was 104.0 nm RMS. It can be converted to MTF of 0.59 using Shannon’s conversion equation [30,31]. This satisfies the MTF requirement (0.55) for fabrication, assembly and alignment. The measured MWIR CTF at the corresponding element of 3-group and 1-element (3G1E) of the target frequency from the USAF image was 0.298. This can be converted to 0.234 using the MTF correlation factor . The result satisfies the MTF value of 0.23 allocated for the MWIR optical system.
3.2 Laboratory thermal test
It is expected that the environmental thermal variation can lead to an image quality that is worse than the designed quality. The image quality compensation method used was to move the focal plane (i.e. the detector) for the EO module, and to move the relay lens group for the MWIR module. In this athermal test, the compensators were moved to the optimal position in order to compensate for the decline in imaging performance caused by the change in the equipment operation temperature.
Figure 7 shows the simulated and measured displacements for the compensator. The maximum difference between the simulated and measured values was 0.18 mm in the EO module and 0.10 mm in the MWIR module. 0.18 mm for the EO module is larger than the EO focal depth ( ± 0.05 mm), but we note that it can be easily adjusted within the available movement range of the compensator ( ± 2.5 mm). 0.1 mm for the MWIR module is smaller than the MWIR focal depth ( ± 0.2 mm), indicating that it is very stable and does not need any adjustment in movement of the MWIR relay lens group within the range of temperature change studied here.
3.3 Simulation and testing for vibration
The FEA predicted and experimental acceleration response to the input vibration to the gimbal and the camera is plotted against the frequency in Fig. 8(a) . It is clear that they agree well with each other over the frequency range of up to around 110 Hz, except for the experimental noise consisting of resonances from both damper and gimbal structures. The FEA analysis also predicts the M1 tilt based on the vibration transmitted to the camera from gimbal, as shown in Fig. 8(b). The maximum M1 slope variation of ± 0.0015 degrees is smaller than the tilt tolerance of M1 and M2 ( ± 0.006 deg.) by about 1/4. The magnitude of M2 slope variation is almost the same as that of M1. Such results support the premise that the camera response to the input vibration is satisfactory. Figure 8(c) is the plot that shows the LOS change caused by M1 and M2 tilts over the duration of vibration. The LOS change is smaller than the maximum ± 40 µrad (micro radian), and is much smaller than the LOS requirement of the entire system including camera, gimbal, and shroud. We can assert that the resulting LOS change under the input vibration would not cause significant performance degradation effects on the whole system, including the camera and the shroud.
3.4 MTF measured in field
As shown in Fig. 9 , there are two day time images (Fig. 9(a) for EO and 9(b) for MWIR) taken from 6 km in distance using custom built EO/MWIR detectors with 8/15 microns in pitch size and 4.19/1.04 Mpixels in pixel number. The measured MTF from the images was 0.321 ± 0.05 (EO) and 0.203 ± 0.06 (MWIR) at the target spatial frequencies as in Section 2.1. These results are slightly better than the requirements, approaching very close to the target performance as summarized in Table 4 . It also shows that the optical design MTF (0.65 for EO and 0.42 for MWIR) in Fig. 3 and Table 1 in Section 2.1 is degraded to the measured MTF of 0.59 (EO) and 0.234 (MWIR) in Section 3.1.
4. Results and discussion
In this study, we report a new dual band telecentric LOROP camera (LC11) that has the smallest volume and weight of all comparable cameras. Table 5 compares our LC11 with other existing cameras [6, 12, 33] in terms of volume and weight. It is clear that the LC11 volume and weight are less than half the largest, and around 47-58% of the averaged volume and mass among those listed in Table 5. Initial laboratory and field performance measurements indicate that LC11 satisfies all the performance targets in Table 1 and 4. We emphasize that it has unique characteristics, including i) compactness in volume and weight, ii) modular design and construction enabling easy maintenance and iii) relaxed tolerances through the first incorporation of telecentricity into a LOROP camera.
We would like to express our thanks to Lee, Eun-jong of Samsung Thales for his support in mechanics, as well as to Sergey Marchuk, Sang-Young Park of Samsung Thales, Hooshik Kim in Vieworks, Il-Kwon Moon, Ho-Soon Yang, Hyuk-Gyo Lee in Korea Research Institute of Standards and Science for useful technical discussions. The manuscript preparation is supported, in part, by NRF research grants 2011-0030-055 and 2011-0020-672.
References and links
1. V. Petrushevsky, “High-resolution long-range oblique IR imaging from an airborne platform,” Proc. SPIE 6395, 1–9 (2006).
2. A. G. Lareau, “Electro-optical imaging array with motion compensation,” U.S. patent 5,155,597 (Oct. 13, 1992).
3. A. G. Lareau, “Electro-optical imaging array with motion compensation,” Proc. SPIE 2023, 65–79 (1993). [CrossRef]
4. A. G. Lareau, “Electro-optical imaging detector array for a moving vehicle which includes two axis image motion compensation and transfers pixels in row directions and column directions,” U.S. patent 5,798,786 (Aug. 25, 1998).
5. R. G. Sementelli, “EO/IR dual-band reconnaissance system DB-110,” Proc. SPIE 2555, 222–231 (1995). [CrossRef]
6. R. N. Lane and J. K. Delaney, “DB-110 performance update,” Proc. SPIE 3431, 108–118 (1998). [CrossRef]
7. K. Riehl Jr., “RAPTOR (DB-110) reconnaissance system: in operation,” Proc. SPIE 4824, 1–12 (2002). [CrossRef]
8. D. Lange, W. Abrams, M. A. Iyengar, R. Lane, and A. Defrietas, “Goodrich DB-110 system: multiband operation today and tomorrow,” Proc. SPIE 5109, 22–36 (2003). [CrossRef]
9. Goodrich Co, Ltd., “Goodrich DB-110 aerial reconnaissance pod provides real-time critical intelligence for defense missions,” http://www.goodrich.com/gr-ext-templating/images/Goodrich%20Content/Enterprise%20Content/Market%20Capabilities/Defense%20and%20Space/51440_DB-110_Reconnaissance.pdf.
10. A. Bodkin, A. Sheinis, and J. McCann, “Compact multi-band (VIS/IR) zoom imager for high resolution long range surveillance,” Proc. SPIE 5783, 816–826 (2005). [CrossRef]
11. J. M. Topaz, D. Freiman, and I. Porat, “Dual-wavelength camera for long-range reconnaissance platforms,” Proc. SPIE 4820, 728–735 (2003). [CrossRef]
12. D. M. Stuart, “Sensor design for unmanned aerial vehicles,” Proc. IEEE 3, 285–295 (1997).
13. Opto Engineering, “Telecentric lenses tutorial,” http://www.opto-engineering.com/telecentric-lenses-tutorial.html.
14. M. Watanabe and S. K. Nayar, “Telecentric optics for constant magnification imaging,” Technical report, Department of Computer Science, Columbia University CUCS-026–95, Sept. 1995.
15. R. A. Petrozzo and S. W. Singer, “Telecentric lenses simplify non-contact metrology,” Test & Measurement World Magazine (Oct. 4–9, 2001).
19. T. Arai and K. Yano, “Telecentric zoom lens,” U.S. patent 7,177,090 (Feb. 13, 2007).
20. N. K. Kawasaki and M. A. Oyama, “Telecentric zoom lens,” U.S. Patent 5,764,419 (June 9, 1998).
21. F. Watanabe, “Telecentric projection lens system,” U.S. Patent 5,905,596 (May 18, 1999).
22. M. Tateoka, “Telecentric projection lenses,” U.S. Patent 4,441,792 (Apr. 10, 1984).
24. J.-Y. Han, S. Marchuk, H. Kim, C. Kim, and K. Park, “Imaging EO/IR optical system for long range oblique photography,” Proc. SPIE 8020, 802009, 802009-6 (2011). [CrossRef]
25. K.-W. Park, Y.-S. Shin, C.-W. Kim, and S.-W. Kim, “Airborne frame camera MTF characteristics due to the B.S.M. (Back Scan Mechanism),” in Proceedings of ADD 40th Anniversary Meeting, Y.-S. Kim, ed. (Daejeon Convention Center, Daejeon, Korea, 2010), pp. 122–125.
28. J. C. Leachtenauer and R. G. Driggers, Surveillance and Reconnaissance Imaging Systems Modeling and Performance Prediction (Artech House, 2001), Chap. 10.
29. R. J. Noll, “Zernike polynomials and atmospheric turbulence,” J. Opt. Soc. Am. 66(3), 207–211 (1976). [CrossRef]
30. R. R. Shannon, The Art and Science of Optical Design (Cambridge University Press, 1997), Chap. 4.
31. W. B. King, “Correlation between the relative modulation function and the magnitude of the variance of the wave-aberration difference function,” J. Opt. Soc. Am. 59(6), 692–697 (1969). [CrossRef]
32. J. W. Coltman, “The specification of imaging properties by response to a sine wave input,” J. Opt. Soc. Am. 44(6), 468–469 (1954). [CrossRef]
33. Aerospace Research Information Center, “Recon/optical CA-295 dual-band digital framing camera,” www.aric.or.kr/trend/accessory/content.asp?classify=10&search=&idx=605&page=1.