|
Article information
2022 , Volume 27, ¹ 3, p.6-15
Gromilin G.I., Kosykh V.P., Likhachev A.V., Shakenov A.K.
Modelling of a random spatially inhomogeneous dynamic background on halftone images
Purpose. In the problems of detecting and tracking moving objects in a sequence of images, the background is often represented by a collection of randomly located areas of different textures. Matched filtering in this case does not allow reliably detecting objects with amplitude of the order of the background level. The method of inter-frame processing developed earlier by the authors provides background suppression, but does not completely solve the problem. Its development requires a model of a spatially non-stationary background. The paper addresses this issue. Methodology. The background model is formed as an applicative mixture of components that are realizations of spatially stationary random processes. Autocorrelation functions (ACF) are a Gaussian or exponential ellipse modulated by a harmonic function. Let 𝑁 (𝑥, 𝑦) stands as uncorrelated, centered, normally distributed noise, and 𝐾(𝜏𝑥, 𝜏𝑦) to be a given ACF. The signal obtained by filtering of 𝑁 (𝑥, 𝑦) by a linear filter with a frequency response equal to the square root of the Fourier transform of 𝐾(𝜏𝑥, 𝜏𝑦) has its ACF 𝐾(𝜏𝑥, 𝜏𝑦). To model a background, one needs having at least two such signals with different ACFs. Their random location on the frame is provided by a binary mask, which is also constructed by filtering a random field. The model provides the formation of images of a dynamic background, for which an independent transformation of its components is carried out by means of an affine transformation. Findings. The model is implemented in Matlab and C++ software environments. The images obtained are visually similar to aerial photographs of the Earth’s surface through the cloud layer. The standard deviation between the ACF of the background component and its estimate for ten thousand implementations was 4 %. On a personal computer, the generation of an image of one component with a size of 1024×1024 pixels is performed in 0.065 seconds. Originality. A simulation model of a spatially non-stationary random dynamic multicomponent background is proposed. Its main purpose is to test algorithms for detecting and tracking small-sized, low-contrast objects in a changing environment. The computational experiment showed the adequacy of the model and the high speed of its execution.
[full text] [link to elibrary.ru]
Keywords: spatially inhomogeneous background, stationary random process, autocorrelation function, linear filter
doi: 10.25743/ICT.2022.27.3.002
Author(s):
Gromilin Gennadiy Ivanovich
Position: General Expert
Office: Institute Automation and Electrometry SB RAS
Address: 630090, Russia, Novosibirsk, 1, Academician Koptyug prosp.,
Phone Office: (383) 330 83 55
E-mail: gromilin@iae.nsk.su
SPIN-code: 9052-3496Kosykh Valeriy Petrovich
PhD.
Position: Head of Laboratory
Office: Institute Automation and Electrometry SB RAS
Address: 630090, Russia, Novosibirsk, 1, Academician Koptyug prosp.,
Phone Office: (383) 330 83 55
E-mail: kosych@iae.nsk.su
SPIN-code: 6144-1243Likhachev Alexey Valerievich
Dr.
Position: Senior Research Scientist
Office: Institute of Automation and Electrometry Siberian Branch of Russian Academy of Science
Address: 630090, Russia, Novosibirsk, Academician Koptug ave. 1
Phone Office: (383) 330 82 43
E-mail: ipm1@iae.nsk.su
SPIN-code: 4283-4592Shakenov Adilbek Koblanovich
PhD.
Position: Junior Research Scientist
Office: Institute of Automation and Electrometry Siberian Branch of Russian Academy of Science
Address: 630090, Russia, Novosibirsk, Academician Koptug ave. 1
Phone Office: (383) 330 83 55
E-mail: Shakenov@iae.nsk.su
SPIN-code: 8583-7717 References:
1. Pratt W.K. Digital image processing. N.Y.: Wiley; 1978: 750.
2. Huang T.S., Yang G.J., Tang G.Y. A fast two-dimensional median filtering algorithm. IEEE Transactions on Acoustics, Speech and Signal Processing. 1979; 27(1):13–18. DOI:10.1109/TASSP.1979.1163188.
3. Tomasi C., Manduchi R. Bilateral filtering for gray and color images. Proceedings of the 1998 IEEE International Conference on Computer Vision. Bombay; 1998: 839–846.
4. Buades A., Morel J.M. A non-local algorithm for image denoising. IEEE Computer Society Conference on Computer Vision and Pattern Recognition. 2005; 2(7):60–65.
5. Likhachev A.V. Modified method for detection small structures in noisy images. Optoelectronics, Instrumentation and Data Processing. 2019; 55(6):580–586. DOI:10.3103/S8756699019060074.
6. Bernstein A.V., Kuleshov A.P. Optimal filtering of a random background in image processing problems. Problems of Information Transmission. 2008; 44(3):233–242. DOI:10.1134/S003294600803006X.
7. Tartakovsky À., Brown J. Adaptive spatial-temporal filtering methods for clutter removal and target tracking. IEEE Transactions-AES. 2008; 44(4):1522–1537. DOI:10.1109/TAES.2008.4667727.
8. Jin L., Hong-Bing J. Improved background prediction algorithm for IR small targets detection. Proceedings îf 2-nd International Congress on Image and Signal Processing (CISP 2009). China, Tianjin; 2009: 3363–3367. DOI:10.1109/CISP.2009.5301700.
9. Muraviev V.S., Muraviev S.I. An adaptive algorithm of aerial object extraction and detection on images. Information and Control Systems. 2011; (5):8–11. (In Russ.)
10. Tartakovsky A.G., Brown A.P., Brown J. Nonstationary EO/IR clutter suppression and dim object tracking. Proceedings of the 2010 Advanced Maui Optical and Space Surveillance Technologies (AMOS) Conference. Maui, Hawaii, September 14–17, 2010. Available at: https://amostech.com/2010-technical-papers (accessed 19.04.2019).
11. Kirichuk V.S., Kosykh V.P., Popov S.A., Sinel’shchikov V.V. Suppression of a quasistationary background in a sequence of images by means of interframe processing. Optoelectronics, Instrumentation and Data Processing. 2014; 50(2):109–117. DOI:10.3103/S8756699014020010.
12. Gromilin G.I., Kosykh V.P., Popov S.A., Streltsov V.A. Suppression of the background with drastic brightness jumps in a sequence of images of dynamic small-size objects. Optoelectronics, Instrumentation and Data Processing. 2019; 55(3):213–221. DOI:10.3103/S8756699019030014.
13. Astafurov V.G., Kur’yanovich K.V., Skorokhodov A.V. A statistical model for describing the texture of cloud cover images from satellite data. Russian Meteorology and Hydrology. 2017; 42(4):248–257. DOI:10.3103/S1068373917040057.
14. Grudin B.N., Plotnikov V.S., Smol’yaninov N.A. Modeling images with specified fractal characteristics. Optoelectronics, Instrumentation and Data Processing. 2010; 46(3):215–221. DOI:10.3103/S8756699010030027.
15. Palagin Yu.I., Fedotov S.V., Shalygin A.S. Mathematical modelling of inhomogeneous random fields and nonstationary processes. Avtomatika i Telemekhanika. 1986; (4):56–67. (In Russ.)
16. Levin B.R. Teoreticheskie osnovy statisticheskoy radiotekhniki [Theoretical foundations of statistical radio engineering]. Moscow: Sovetskoe radio; 1969: 751. (In Russ.)
Bibliography link:
Gromilin G.I., Kosykh V.P., Likhachev A.V., Shakenov A.K. Modelling of a random spatially inhomogeneous dynamic background on halftone images // Computational technologies. 2022. V. 27. ¹ 3. P. 6-15
|
