کاهش نوفه تصاویر نجومی با استفاده از معادلات مشتقات جزئی

نوع مقاله : مقاله پژوهشی‌

نویسندگان

دانشکده مهندسی نقشه‌برداری و اطلاعات مکانی، پردیس دانشکده‌های فنی، دانشگاه تهران، تهران، ایران

چکیده

استفاده از سیستم­های بینایی مبنای نجومی به‌عنوان روشی ارزان و مناسب به‌منظور تعیین مختصات نقاط می­تواند به‌عنوان روشی کمکی و همچنین جایگزین برای سیستم­های تعیین موقعیت جهانی در نظر گرفته شود. علاوه بر آن با استفاده از این سیستم می‌توان حرکت ماهواره­های جاسوسی را نیز رصد نمود. با توجه به استفاده این سیستم­ها از تصاویر رقومی، کیفیت حاصله نقش مهمی در کیفیت خروجی نهایی خواهد داشت. نوفه‌های ایجاد شده در مرحله اخذ تصویر ازجمله اثراتی هستند که باعث مخدوش شدن کیفیت تصویر می­شوند. در این مقاله الگوریتم جدیدی با الهام از روش­های مبتنی بر معادلات نفوذ گرما برای کاهش نوفه تصاویر نجومی ارائه گردیده است. از ویژگی­های مهم این روش جلوگیری از تغییر مکان عوارض در تصاویر یا به‌عبارت‌دیگر حفظ ساختار تصویر است زیرا تشخیص بسیار دقیق مراکز ستارگان تأثیر بسزایی در انطباق آنها با کاتالوگ خواهد داشت. این روش بر این اصل استوار است که روشنایی لبه­های موجود در تصاویر واقعی به‌تدریج تغییر می­یابد. بر این اساس برای حل عددی معادلات نفوذ گرما به‌منظور حفظ لبه­ها، لازم است همسایگی­هایی با عمق بیش از یک پیکسل در نظر گرفته شود. با این کار موقعیت­های واقعی حفظ می­شوند. این الگوریتم بر روی تصاویر مختلف آزمایش‌ شده و عملکرد آن با الگوریتم­های موجود مقایسه شده است. نتایج عددی حاصل از مقایسه چهار معیار بین الگوریتم پیشنهادی با فیلتر میانگین و میانه نشان می­دهد که روش ارائه ‌شده در این مقاله، از عملکرد بهتری برخوردار است.

کلیدواژه‌ها


عنوان مقاله [English]

Nonlinear diffusion noise reduction in astronomical images

نویسندگان [English]

  • Mohammad Ali Sharifi
  • Saeed Farzaneh
  • mona kosary
Department of Surveying and Geomatics Engineering, Faculty of Engineering, University of Tehran, Tehran, Iran
چکیده [English]

From ancient times, celestial bodies have been used by travelers and scientists for positioning and routing. By developing sciences, it was found that the celestial bodies form an accurate inertial system to use in navigation applications. In this system, each star is considered as a reference point in determining reference coordinate frame of the system. Due to the visibility of the satellite motion trace and the fundamental need to determine and modify satellites’ orbital parameters as well as to identify and locate espial satellites, determining the positional parameters of the satellite is also one of the modern and important applications of vision-based astronomical systems. In the modern vision-based astronomical systems, data collection is done using charge-coupled device (CCD) array. During the process of light collision to the surface of the CCD and then reading and measuring the number of photoelectrons as well as converting them to the digital numbers to store them as grey degree in each pixel, the smallest mistakes that result in lost or added electrons on each pixel can lead to distortion and noise in the image. The process of noise elimination should not only eliminate or reduce the noise but also avoid blurring the image and removing or relocating the edges of the image. To determine the primary orbit of the satellite using an optical method, the streak of the satellite must be extracted accurately because the misdiagnosis of the beginning and end points of the streak directly affects the accuracy of the determined orbit. Therefore, we need to find noise elimination methods that impose the lowest possible effects to the key complications of the astronomical images such as star and satellite streak. In this study, it is attempted to eliminate the noises using diffusion equation and solving it numerically. On the other hand, to identify the accurate position of the edges, the gradient is calculated by through convoluting the main image by the Gaussian filter. In this study, a numerical method is used to solve diffusion equation. Heat diffusion equation is an iteration-based method. It is obvious that the more the paces and iterations in the equation, the smoother the image. This factor must be chosen such that the image brightness does not exceed the main range. For this purpose, the noise must be eliminated from the image by choosing appropriate number of iterations. In this research, the structural similarity index (SSI) is used to select the optimum number of iterations. As a result, in this research, it is attempted to use noise elimination methods that impose the lowest changes to the existing satellite’s streak.

کلیدواژه‌ها [English]

  • star centroiding
  • nonlinear diffusion
  • Noise Reduction
  • satellite streak
Altman, D. and Bland, J., 1994, Diagnostic tests. 1: Sensitivity and specificity: British Medical, 308(6943), 1552.
Arias-Castro, E. and Donoho, D. L., 2009, Does median filtering truly preserve edges better than linear filtering?: The Annals of Statistics, 1172-1206.
Ben Said, A., Hadjidj, R., Eddine Melkemi, K., and Foufou, S., 2016, Multispectral image denoising with optimized vector non-local mean filter: Digital Signal Processing, 58, 115-126.
Black, M. J., Sapiro, G., Marimont, D. H., and Heeger, D., 1998, Robust anisotropic diffusion: IEEE Transactions on image processing, 7(3), 421-432.
Buades, A., Coll, B., and Morel, J. M., 2005, A non-local algorithm for image denoising: IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).
Buades, A., and Morel, J. M., 2005, A review of image denoising algorithms, with a new one: Multiscale Modeling and Simulation, 4(2), 490-530.
Dong, W. and Ding, H., 2016, Full frequency de-noising method based on wavelet decomposition and noise-type detection: Neurocomputin, 214, 902-909.
Erdem, E., 2013, nonlinear diffusion, Hacettepe University, http://web.cs.hacettepe.edu.tr/~erkut/bil717.s13/w03-nonlineardif.pdf.
Gerig, G., Kubler, O., Kikinis, R., Jolesz, F. A., 1992, Nonlinear anisotropic filtering of MRI data: IEEE Transactions on medical imaging, 11(2), 221-232.
Gonzalez, R. C., and Woods, R. E., 2009, Digital Image Processing.
Hu, K., Cheng, Q., and Gao, X., 2016, Wavelet-domain TI Wiener-like filtering for complex MR data denoising: Magnetic Resonance Imaging, 34(8), 1128-1140.
Lim, J. S., 1990, Two-dimensional signal and image processing.
Perreault, S., and Hébert, P., 2007, Median filtering in constant time: IEEE transactions on image processing, 16(9), 2389-2394.
Perona, P., and Malik, J., 1990, Scale-space and edge detection using anisotropic diffusio: IEEE Transactions on pattern analysis and machine intelligence, 12(7), 629-639.
Rajan, J., Kannan, K., and Kaimal, M., 2008, An improved hybrid model for molecular image denoising: Journal of Mathematical Imaging and Vision, 31(1), 73-79.
Farzaneh, S., Sharifi, M. A., and Kosary, M., 2017, Automatic satellite streaks detection in astronomical images: Journal of the Earth and Space Physics, 43(3), 473-487.
Tomasi, C., and Manduchi, R., 1998, Bilateral filtering for gray and color images, In Computer Vision,Sixth International Conference on, 839-846, IEEE.
Tsiotsios, C., and Petrou, M., 2013, On the choice of the parameters for anisotropic diffusion in image processing: Pattern Recognition, 46(5), 1369-1381.
Wang, Z., and Sheikh, H. R., et al., 2004, Image quality assessment: from error visibility to structural similarity: IEEE transactions on image processing, 13(4), 600-612.
Weeratunga, S. K., and Kamath, C., 2002, PDE-based nonlinear diffusion techniques for denoising scientific and industrial image: International Society for Optics and Photonics.
Weeratunga, S. K., and Kamath, C., 2003, Comparison of PDE-based non-linear anistropic diffusion techniques for image denoising: International Society for Optics and Photonics.
Weickert, J., 1997, A review of nonlinear diffusion filtering: in International Conference on Scale-Space Theories in Computer Vision, Springer.
Wong, A., Mishra, A., Bizheva, K., and Clausi, D. A., 2010, General Bayesian estimation for speckle noise reduction in optical coherence tomography retinal imagery: Optics express, 18(8), 8338-8352.