A methods for synthesizing images of target environment taking into account the characteristics of a thermal electro-optical imaging system


Аuthors

Krasnov A. M.1*, Shashkov S. N.2, Rumyantsev A. V.3

1. Closed Joint Stock Company «Technological Park of Cosmonautics «LINKOS»», Moscow, Shcherbinka, Russia
2. State Research Institute of aviation systems, Moscow, Russia
3. Сenter (control of integrated safety and security) the MoD RF , Moscow, Russia

*e-mail: a_krasnov@inbox.ru

Abstract

This article discusses a generalized training scheme for a neural network designed for object detection and recognition, as well as a method for synthesizing background-target images based on the characteristics of a imaging electro-optical system thermal system (EOS). This method includes the following stages: initial data preparation; energy calculation and synthesis of a set of initial images of objects and background environments; and synthesis of images of the background-target environment. The relevance of this material stems from the need to create training and test image sets for the neural network, taking into account EOS characteristics such as the signal transfer function, 3D noise models, and modulation transfer function. The system intensity transfer function (SiTF) characterizes the EOS sensitivity, the 3D noise model represents information about the EOS's temporal and spatial noise, and the modulation transfer function (MTF) is the key characteristic for measuring the EOS's resolution. The most fundamental stage of the methodology - initial data preparation - is examined in detail. The entire data set is divided into three groups: initial data on the observed object and the underlying surface (background); initial data on the state of the atmosphere (external conditions); and the characteristics and parameters of the EOS. The initial data preparation stage concludes with the specification of the target background environment and external conditions. Then, the synthesis of the set of initial object images (background environment) and the synthesis of target background environment images are described. The example presents the results obtained during each stage and demonstrates the initial and synthesized target background images. Following this, the necessary support is determined, including software, measuring instruments and workstations, and logistics. Next, the methods for assessing the conformity of the synthesized target background images with the images generated by the EOS is discussed.

Keywords:

synthesized background-target images, electro-optical system, signal transfer function, 3D noise model, modulation transfer function, neural network

References

  1. Corey D. Packard, Mark D. Klein, Timothy S. Viola, David C. Bell, Peter L. Rynes, "Automated simulation-generated EO/IR image library for arti-ficial intelligence applications," Proc. SPIE 11406, Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XXXI, 1140606 (23 April 2020); doi: 10.1117/12.2558843. 
  2. Ronald G. Driggers, Melvin H. Friedman, John W. Devitt, Orges Furxhi, Anjali Singh. Introduction to Infrared and Electro-Optical Systems - Third Edition. Artech House, 2022. 
  3. Alijafar Mousivand, Wout Verhoef, Massimo Menenti, Ben Gorte “Modeling Top of Atmosphere Radiance over Heterogeneous Non-Lambertian Rugged Terrain”. Remote Sens. 2015, 7, 8019-8044; doi:10.3390/rs70608019. 
  4. Gerald C. Holst. Electro-Optical Imaging System Performance - Fifth edition. JCD Publishing and SPIE Press, 2008. 
  5. A. A. Chertoriyskiy. Methods and technical means of measuring optical radiation parameters. The training manual. Ulyanovsk, ULSTU, 2020. 
  6. Dataset Mid-IR Spectrometer ATP8250, www.optosky.com , Opto-sky Photonics Inc., 2015. 
  7. Khismatov I.F. Methodology of reproducing meteorological conditions in the simulation of aviation optoelectronic systems // Proceedings of the MAI. Issue No. 108. https://doi.org/10.34759/trd-2019-108-18 . 
  8. Czrzanowski K. Testing thermal imagers. Practical guidebook. // Mili-tary University of Technology, 00-908 Warsaw, Poland, 2010. 
  9. Krasnov A.M., Tregubenkov S.Yu., Rumyantsev A.V., Khismatov R.F., Shashkov S.N. Evaluation of the effectiveness of optoelectronic systems using an analytical model. Functions of threshold contrast and modulation transmission // Information-measuring and control systems. 2021. Vol. 19. No. 1. pp. 45-64. DOI: 10.18127/j20700814-202101-04 
  10. Krasnov A.M., Tregubenkov S.Yu., Rumyantsev A.V., Khismatov R.F., Shashkov S.N. Evaluation of the effectiveness of optoelectronic systems using an analytical model. The noise model of the ECO-Operator system. Proceedings of MAY 2022. No. 122. DOI: 10.34759/trd-2022-122-22 . 
  11. J. A. D’Agostino and C. M. Webb, «Three-dimensional analysis framework and measurement methodology for imaging system noise,» Proc. SPIE 1488, 110 (1991). 
  12. Patrick O’Shea, Stephen Sousk “Practical Issues with 3D-Noise Measurements and Application to Modern Infrared Sensors”. Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XVI, Proceedings of SPIE Vol. 5784, 2005; doi: 10.1117/12.604588. 
  13. Alan Irwin, "Performance prediction from EO system measurements using IRWindows and NV-IPM," Proc. SPIE 10625, Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XXIX, 1062502 (26 April 2018); doi: 10.1117/12.2315492.
  14. Gerald C. Holst. Electro-Optical Imaging System Performance - Sixth edition. JCD Publishing and SPIE Press, 2017. 
  15. Gonzalez R., Woods R., Eddins S. Digital image processing in the MATLAB environment. Moscow: Technosphere, 2006. 
  16. T. Saati. Decision-making. A method for analyzing hierarchies. Moscow, Radio and Communications, 1993. 
  17. Saati T.L. Decision-making with dependencies and feedback: analytical networks. Translated from English/Scientific ed. by A.V. Andreychikov, Moscow: LKI Publishing House, 2008, 360 s.
  18. N.S. Kasimov, E.I. Golubeva, I.K. Lurie, M.V. Zimin, T.E. Samsonov, O.V. Tutubalina, U.G. Ris, A.I. Mikheeva, A.R. Alyautdinov. Library of spectral characteristics of geographical objects in the structure of the Lomonosov Moscow State University Geoportal. Bulletin of the Moscow University. Series 5. GEOGRAPHY. 2015. № 5
  19. GOST R57700.37-2021. Computer models and modeling. Digital counterparts of products. General provisions
  20. GOST R 57700.24. Computer models and modeling. Validation basis.
  21. GOST R 57700.22. Computer models and modeling. Classification
  22. GOST R 57412. Computer models in the processes of product development, production and operation. General provisions.


Download

mai.ru — informational site MAI

Copyright © 2000-2025 by MAI

 Вход