Deep learning enables parallel camera with enhanced- resolution and computational zoom imaging
doi: 10.1186/s43074-023-00095-3
Deep learning enables parallel camera with enhanced- resolution and computational zoom imaging
-
Abstract:
High performance imaging in parallel cameras is a worldwide challenge in computational optics studies. However, the existing solutions are suffering from a fundamental contradiction between the field of view (FOV), resolution and bandwidth, in which system speed and FOV decrease as system scale increases. Inspired by the compound eyes of mantis shrimp and zoom cameras, here we break these bottlenecks by proposing a deep learning-based parallel (DLBP) camera, with an 8-μrad instantaneous FOV and 4 × computational zoom at 30 frames per second. Using the DLBP camera, the snapshot of 30-MPs images is captured at 30 fps, leading to orders-of-magnitude reductions in system complexity and costs. Instead of directly capturing photography with large scale, our interactive-zoom platform operates to enhance resolution using deep learning. The proposed end-to-end model mainly consists of multiple convolution layers, attention layers and deconvolution layer, which preserves more detailed information that the image reconstructs in real time compared with the famous super-resolution methods, and it can be applied to any similar system without any modification. Benefiting from computational zoom without any additional drive and optical component, the DLBP camera provides unprecedented-competitive advantages in improving zoom response time (~ 100 ×) over the comparison systems. Herein, with the experimental system described in this work, the DLBP camera provides a novel strategy to solve the inherent contradiction among FOV, resolution and bandwidth.
-
Key words:
- Imaging system /
- Parallel camera /
- Compound eye /
- Computational imaging /
- Deep learning /
- Super resolution /
- Interactive panorama
-
[1] Ditchburn RW. Information and control in the visual system. Nature. 1963;198:630. [2] Brady DJ, et al. Multiscale gigapixel photography. Nature. 2012;486:386–9. [3] Brady DJ, et al. Characterization of the AWARE 40 wide-field-of-view visible imager. Optica. 2015;2(12):1086. [4] Brady DJ, et al. Parallel cameras Optica. 2018;5(2):127–37. [5] Fan JT, et al. Video-rate imaging of biological dynamics at centimetre scale and micrometre resolution. Nat Photonics. 2019;13:809–16. [6] Strogatz SH. Exploring complex networks. Nature. 2001;410:268–76. [7] Kittle DS, et al. A testbed for wide-field, high-resolution, gigapixel-class cameras. Rev Sci Instrum. 2013;84: 053107. [8] Park HJ, et al. Structural and functional brain networks: from connections to cognition. Science. 2013;342:1238411. [9] Bullmore E, et al. Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci. 2009;10:186–98. [10] Wilburn B, et al. High performance imaging using large camera arrays. ACM Trans Graph. 2005;24:765–76. [11] Lynn CW, et al. The physics of brain network structure, function and control. Nat Rev Phys. 2019;1:318–32. [12] Seshadrinathan K, et al. High dynamic range imaging using camera arrays. 2017 IEEE International Conference on Image Processing (ICIP). IEEE; 2017. p. 725-9. [13] Zhang Y, et al. Multi-focus light-field microscopy for high-speed large-volume imaging. PhotoniX. 2022;3:1–20. [14] Wu JC, et al. An integrated imaging sensor for aberration-corrected 3D photography. Nature. 2022;612:62–71. [15] Cossairt, et al. Scaling law for computational imaging using spherical optics. JOSA A. 2011;28:2540–53. [16] Jeong KH, et al. Biologically inspired artificial compound eyes. Science. 2006;312(5773):557–61. [17] Zhu L, et al. Miniaturising artificial compound eyes based on advanced micronanofabrication techniques. Light: Adv Manuf. 2021;2(1):84–100. [18] Cao XW, et al. Single-pulse writing of a concave microlens array. Opt Lett. 2018;43:831–4. [19] Tanida J, et al. Color imaging with an integrated compound imaging system. Opt Express. 2003;11:2109–17. [20] Wu D, et al. High numerical aperture microlens arrays of close packing. Appl Phys Lett. 2010;97(3):031109. [21] Chan EP, et al. Fabricating microlens arrays by surface wrinkling. Adv Mater. 2006;18:3238–42. [22] Song YM, et al. Digital cameras with designs inspired by the arthropod eye. Nature. 2013;497:95–9. [23] Cheng Y, et al. Review of state-of-the-art artificial compound eye imaging systems. Bioinspir Biomim. 2019;14(3): 031002. [24] Park SH, et al. Subregional slicing method to increase three-dimensional nano-fabrication efficiency in two-photon polymerization. Appl Phys Lett. 2005;87:154108. [25] Kirschfeld K. The resolution of lens and compound eyes. Neural principles in vision. 1976. p. 354–70. [26] Cossairt OS, et al. Gigapixel computational imaging. 2011 IEEE International Conference on Computational Photography (ICCP). 2011. p. 1–8. [27] Liu SB, et al. Real-time and ultrahigh accuracy image synthesis algorithm for full field of view imaging system. Sci Rep. 2020;10(1):12389. [28] Perazzi F, et al. Panoramic video from unstructured camera arrays. Computer Graph Forum. 2015;34:57–68. [29] Dai QH, et al. A modular hierarchical array camera. Light Sci Appl. 2021;10(1):1–9. [30] Afshari H, et al. A spherical multi-camera system with real-time omnidirectional video acquisition capability. IEEE T Consum Electr. 2012;58:1110–8. [31] Cohen MF, et al. Capturing and viewing gigapixel images. ACM Trans. Graph. 2007;26(3): 93–es. [32] Gigapan time machine. (2016). [Online]. Available: http://timemachine.cmucreatelab.org. [33] Ivezić Ž, et al. LSST: from science drivers to reference design and anticipated data products. American Astronomical Society Meeting. 2009;213:460–03. [34] Marks DL, et al. Characterization of the AWARE 10 two-gigapixel wide-field-of-view visible imager. Appl Opt. 2014;53(13):C54–63. [35] Hou C, et al. Ultra slim optical zoom system using Alvarez freeform lenses. IEEE Photonics J. 2019;11(6):1–10. [36] Zou Y, et al. Ultra-compact optical zoom endoscope using solid tunable lenses. Opt Express. 2017;25(17):20675–88. [37] Savidis N, et al. Nonmechanical zoom system through pressure-controlled tunable fluidic lenses. Appl Opt. 2013;52(12):2858–65. [38] Zhang DY, et al. Fluidic adaptive zoom lens with high zoom ratio and widely tunable field of view. Opt Commun. 2005;249(1–3):175–82. [39] Cira NJ, et al. Vapour-mediated sensing and motility in two-component droplets. Nature. 2015;519(7544):446–50. [40] Nie J, et al. Self-powered microfluidic transport system based on triboelectric nanogenerator and electrowetting technique. ACS Nano. 2018;12:1491–9. [41] Lee J, et al. Multifunctional liquid lens for variable focus and aperture. Sensor Actuat A-Phys. 2019;287:177–84. [42] Li YL, et al. Tunable liquid crystal grating based holographic 3D display system with wide viewing angle and large size. Light Sci Appl. 2022;11(1):1–10. [43] Jamali A, et al. Large area liquid crystal lenses for correction of presbyopia. Opt Express. 2020;28(23):33982–93. [44] Chu F, et al. Four-mode 2D/3D switchable display with a 1D/2D convertible liquid crystal lens array. Opt Express. 2021;29(23):37464–75. [45] Kuiper S, et al. Variable-focus liquid lens for miniature cameras. Appl Phys Lett. 2004;85(7):1128–30. [46] Son HM, et al. Tunable-focus liquid lens system controlled by antagonistic winding-type SMA actuator. Opt Express. 2009;17(16):14339–50. [47] Lin YH, et al. An electrically tunable optical zoom system using two composite liquid crystal lenses with a large zoom ratio. Opt Express. 2011;19(5):4714–21. [48] Lin HC, et al. A holographic projection system with an electrically tuning and continuously adjustable optical zoom. Opt Express. 2012;20(25):27222–9. [49] Cheng J, et al. CUDA by example: an introduction to general-purpose GPU programming. Scalable Computing: Practice and Experience, 2010;11(4):401. [50] Xing W, et al. Fast pedestrian detection based on haar pre-detection[J]. International Journal of Computer and Communication Engineering. 2012;1(3):207. [51] Henriques JF, et al. High-speed tracking with kernelized correlation filters. IEEE Trans Pattern Anal Mach Intell. 2015;37:583–96. [52] Rublee E, et al. ORB: an efficient alternative to SIFT or SURF. 2011 IEEE International Conference on Computer Vision. 2011. p. 2564–71. [53] Lowe DG. Distinctive image features from scale-invariant keypoints. Int J Comput Vision. 2004;60:91–110. [54] Song ZL, Zhang JP. Remote Sensing Image Registration Based on Retrofitted SURF Algorithm and Trajectories Generated From Lissajous Figures. IEEE GEOSCI REMOTE S. 2010;7:491–5. [55] Fischler MA, et al. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. Commun ACM. 1981;24:381–95. [56] Sanders J, et al. CUDA by example: an introduction to general-purpose GPU programming. Addison-Wesley Professional. 2010. [57] Lai WS, et al. Deep Laplacian pyramid networks for fast and accurate super-resolution. CVPR. 2017. p. 624–32. [58] Park SH, et al. Flexible style image super-resolution using conditional objective. IEEE Access. 2022;10:9774–92. [59] Lim B, et al. Enhanced deep residual networks for single image super-resolution. IEEE Conf. Comput. Vis. Pattern Recognit. 2017. p. 136–44. [60] Chen C, et al. Camera lens super-resolution. IEEE Conf. Comput. Vis. Pattern Recognit. 2019. p. 1652–60. -
下载:
点击查看大图
图(1)
计量
- 文章访问数: 252
- HTML全文浏览量: 7
- PDF下载量: 17
- 被引次数: 0
