Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
166 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

PI-AstroDeconv: A Physics-Informed Unsupervised Learning Method for Astronomical Image Deconvolution (2403.01692v1)

Published 4 Mar 2024 in astro-ph.IM, astro-ph.GA, cs.CV, and eess.IV

Abstract: In the imaging process of an astronomical telescope, the deconvolution of its beam or Point Spread Function (PSF) is a crucial task. However, deconvolution presents a classical and challenging inverse computation problem. In scenarios where the beam or PSF is complex or inaccurately measured, such as in interferometric arrays and certain radio telescopes, the resultant blurry images are often challenging to interpret visually or analyze using traditional physical detection methods. We argue that traditional methods frequently lack specific prior knowledge, thereby leading to suboptimal performance. To address this issue and achieve image deconvolution and reconstruction, we propose an unsupervised network architecture that incorporates prior physical information. The network adopts an encoder-decoder structure while leveraging the telescope's PSF as prior knowledge. During network training, we introduced accelerated Fast Fourier Transform (FFT) convolution to enable efficient processing of high-resolution input images and PSFs. We explored various classic regression networks, including autoencoder (AE) and U-Net, and conducted a comprehensive performance evaluation through comparative analysis.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (55)
  1. Effects of heat sources in the telescope beam on astronomical image quality. In Advanced Technology Optical Telescopes V, volume 2199, pages 478–480. SPIE, 1994.
  2. Analysing the impact of far-out side-lobes on the imaging performance of the ska-low telescope. Monthly Notices of the Royal Astronomical Society, page stw2814, 2016.
  3. Eliminating Primary Beam Effect in Foreground Subtraction of Neutral Hydrogen Intensity Mapping Survey with Deep Learning. Astrophys. J., 934(1):83, 2022. doi:10.3847/1538-4357/ac7a34.
  4. A Covington and N Broten. An interferometer for radio astronomy with a single-lobed radiation pattern. IRE Transactions on Antennas and Propagation, 5(3):247–255, 1957.
  5. David Woody. Radio interferometer array point spread functions i. theory and statistics. ALMA Memo Series, 389, 2001a.
  6. David Woody. Radio interferometer array point spread functions ii. evaluation and optimization. ALMA Memo Series, 390, 2001b.
  7. Neal Jackson. Principles of interferometry. Jets from Young Stars II: Clues from High Angular Resolution Observations, pages 193–218, 2008.
  8. Reduction of sidelobe and speckle artifacts in microwave imaging: The clean technique. IEEE Transactions on Antennas and Propagation, 36(4):543–556, 1988.
  9. K. Rohlfs and T. L. Wilson. Interferometers and Aperture Synthesis, pages 192–226. Springer Berlin Heidelberg, Berlin, Heidelberg, 1996. ISBN 978-3-662-03266-4. doi:10.1007/978-3-662-03266-4_8. URL https://doi.org/10.1007/978-3-662-03266-4_8.
  10. Blind image deconvolution. IEEE signal processing magazine, 13(3):43–64, 1996.
  11. Fast image deconvolution using hyper-laplacian priors. Advances in neural information processing systems, 22, 2009.
  12. Motion blur kernel estimation via salient edges and low rank prior. In 2014 IEEE International Conference on Multimedia and Expo (ICME), pages 1–6. IEEE, 2014.
  13. Image deblurring via enhanced low-rank prior. IEEE Transactions on Image Processing, 25(7):3426–3437, 2016.
  14. l⁢_⁢0𝑙_0l\_0italic_l _ 0-regularized intensity and gradient prior for deblurring text images and beyond. IEEE transactions on pattern analysis and machine intelligence, 39(2):342–355, 2016.
  15. Underwater image restoration by red-dark channel prior and point spread function deconvolution. In 2015 IEEE international conference on signal and image processing applications (ICSIPA), pages 110–115. IEEE, 2015.
  16. Deblurring images via dark channel prior. IEEE transactions on pattern analysis and machine intelligence, 40(10):2315–2328, 2017.
  17. Fast blind deconvolution using a deeper sparse patch-wise maximum gradient prior. Signal Processing: Image Communication, 90:116050, 2021.
  18. Deep convolutional neural network for image deconvolution. Advances in neural information processing systems, 27, 2014.
  19. A machine learning approach for non-blind image deconvolution. In 2013 IEEE Conference on Computer Vision and Pattern Recognition, pages 1067–1074, 2013. doi:10.1109/CVPR.2013.142.
  20. Deep multi-scale convolutional neural network for dynamic scene deblurring. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3883–3891, 2017.
  21. Scale-recurrent network for deep image deblurring. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 8174–8182, 2018.
  22. Deep wiener deconvolution: Wiener meets deep learning for image deblurring. Advances in Neural Information Processing Systems, 33:1048–1059, 2020.
  23. Deep learning for fast spatially varying deconvolution. Optica, 9(1):96–99, 2022.
  24. WSClean: an implementation of a fast, generic wide-field imager for radio astronomy. MNRAS, 444(1):606–619, 2014. doi:10.1093/mnras/stu1368.
  25. Casa, common astronomy software applications for radio astronomy. Publications of the Astronomical Society of the Pacific, 134(1041):114501, 2022.
  26. Tools of radio astronomy. Springer Science & Business Media, 2013.
  27. Sven Treitel. Predictive deconvolution-theory and practice. Geophysics, 34(2):155–169, 1969.
  28. Graph-based blind image deblurring from a single photograph. IEEE Transactions on Image Processing, 28(3):1404–1418, 2019. doi:10.1109/TIP.2018.2874290.
  29. Blind deblurring for saturated images. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 6304–6312, 2021. doi:10.1109/CVPR46437.2021.00624.
  30. Deep learning for handling kernel/model uncertainty in image deconvolution. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 2385–2394, 2020. doi:10.1109/CVPR42600.2020.00246.
  31. Neural blind deconvolution using deep priors. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 3338–3347, 2020. doi:10.1109/CVPR42600.2020.00340.
  32. Learning representations by back-propagating errors. nature, 323(6088):533–536, 1986.
  33. U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III 18, pages 234–241. Springer, 2015.
  34. Generative adversarial networks: An overview. IEEE signal processing magazine, 35(1):53–65, 2018.
  35. Rene Racine. The telescope point spread function. Publications of the Astronomical Society of the Pacific, 108(726):699, 1996.
  36. G Dougherty and Z Kawaf. The point spread function revisited: image restoration using 2-d deconvolution. Radiography, 7(4):255–262, 2001.
  37. Tim J Cornwell. Multiscale clean deconvolution of radio synthesis images. IEEE Journal of selected topics in signal processing, 2(5):793–801, 2008.
  38. Deep image deblurring: A survey. International Journal of Computer Vision, 130(9):2103–2130, 2022.
  39. Boualem Boashash. Time-frequency signal analysis and processing: a comprehensive reference. Academic press, 2015.
  40. Pierre Connes. Astronomical fourier spectroscopy. Annual review of Astronomy and Astrophysics, 8(1):209–230, 1970.
  41. Deconvolution in astronomy: A review. Publications of the Astronomical Society of the Pacific, 114(800):1051, 2002.
  42. Ronald N Bracewell. Strip integration in radio astronomy. Australian Journal of Physics, 9(2):198–217, 1956.
  43. Tensorflow: A system for large-scale machine learning. In 12th {normal-{\{{USENIX}normal-}\}} Symposium on Operating Systems Design and Implementation ({normal-{\{{OSDI}normal-}\}} 16), pages 265–283, 2016.
  44. Pytorch: An imperative style, high-performance deep learning library. In Advances in Neural Information Processing Systems 32, pages 8024–8035. Curran Associates, Inc., 2019. URL http://papers.neurips.cc/paper/9015-pytorch-an-imperative-style-high-performance-deep-learning-library.pdf.
  45. Fft and convolution performance in image filtering on gpu. In Tenth International Conference on Information Visualisation (IV’06), pages 609–614. IEEE, 2006.
  46. Jeffrey Hurchalla. A time distributed fft for efficient low latency convolution. In Audio Engineering Society Convention 129. Audio Engineering Society, 2010.
  47. Fast training of convolutional networks through ffts. arXiv preprint arXiv:1312.5851, 2013.
  48. Fast convolutional neural networks with fine-grained ffts. In Proceedings of the ACM international conference on parallel architectures and compilation techniques, pages 255–265, 2020.
  49. A comprehensive survey of loss functions in machine learning. Annals of Data Science, pages 1–26, 2020.
  50. Lawrence G Burriesci. Nircam instrument overview. In Cryogenic Optical Systems and Instruments XI, volume 5904, pages 21–29. SPIE, 2005.
  51. The near-infrared camera (nircam) for the james webb space telescope (jwst). In Optical, Infrared, and Millimeter Space Telescopes, volume 5487, pages 628–634. SPIE, 2004.
  52. Adam: A method for stochastic optimization. CoRR, abs/1412.6980, 2014. URL https://api.semanticscholar.org/CorpusID:6628106.
  53. Fast in space: Considerations for a multibeam, multipurpose survey using china’s 500-m aperture spherical radio telescope (fast). IEEE Microwave Magazine, 19(3):112–119, 2018. doi:10.1109/MMM.2018.2802178.
  54. The square kilometre array. Proceedings of the IEEE, 97(8):1482–1496, 2009. doi:10.1109/JPROC.2009.2021005.
  55. Hu Zhan. An overview of the chinese space station optical survey. 42nd COSPAR Scientific Assembly, 42:E1–16, 2018.
Citations (1)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now