M.Kh. Aksayitov1, E.V. Egorova2, A.N. Ribakov3

1 JSC Concern Granit-Electron (St. Petersburg, Russia)
2 MIREA – Russian Technological University (RTU MIREA) (Moscow, Russia)
3 Russian Research Institute of Automation n. a. N.L. Dukhov (Moscow, Russia)
2 calipso575@gmail.com

The results of Russian theoretical and experimental studies characterize the main areas of research in the field of detection and recognition of various radar objects, while the main research tool for most works is the detection and development of promising mathematical models of objects and modeling of secondary emissions for their recognition, which allows in some cases to obtain additional information about these objects. The problem of detection, parameter assessment and classification of spatially distributed targets remains decisive in the development and creation of modern radar systems for monitoring the space, air and ground situation by modern and promising radar means. The use of a probabilistic procedure for recognizing and identifying a spatial object from single or spatial images requires the definition of interrelated main stages of recognition, taking into account the specificity and scientific and technical complexity of the task. It should be taken into account that in many cases the operation of identification algorithms can be tested by means of their modeling on a computer without using natural information for these purposes. The complexity of conducting the analysis and formalization of the methodology for determining features based on obtaining expressions of the vector of sufficient statistics in each specific observation situation and transforming the said vector in such a way as to preserve all of its parameters, a priori information about which the classifier has or which can be "trained" as a result of "worker-like" training allows us to confirm the efficiency of the classification performed according to the proposed method for typical spatially distributed targets that differ in size and specific effective scattering area. The conducted assessment of the angle by the orientation of the contour of the spatially distributed target and its mirror image based on the analysis of a real radar target allows us to determine the structure of the optimal algorithm for the most common types of distributions of the one-dimensional probability density of radar image readings. The analysis of classification efficiency, carried out using the proposed method for typical spatially distributed targets differing in size and specific effective scattering area, shows that to ensure satisfactory classification efficiency (SSA > 0.9) at a resolution of about 20–30 m, it is sufficient for targets to differ in size by 25–30% and in total effective scattering area by 3 dB.

Aksayitov M.Kh., Egorova E.V., Ribakov A.N. Optimal algorithm for determining the angle of a spatially distributed target. Science Intensive Technologies. 2024. V. 25. № 6. P. 26−33. DOI: https://doi.org/10.18127/j19998465-202406-04 (in Russian)

1. Anderson T. Vvedenie v mnogomernyj statisticheskij analiz: Per. s angl. B.V. Gnedenko. M.: Fizmatgiz. 1963. S. 500 (in Russian).
2. Harkevich A.A. Izbrannye trudy. V 3 t. Teoriya informacii. Opoznavanie obrazov. T. 3. M.: Nauka. 1973. S. 524 (in Russian).
3. Sebastian G.S. Processy prinyatiya reshenij pri raspoznavanii obrazov: Per. s angl. Ivah-nenko. Kiev: Tekhnika. 1965. S. 152 (in Russian).
4. Bongard M.M. Problema uznavaniya. M.: Nauka. 1967. S. 320 (in Russian).
5. Vapnik V.N., Chervonenkis A.Ya. Teoriya raspoznavaniya obrazov (statisticheskie problemy obu-cheniya). M.: Nauka. 1974. S. 616 (in Russian).
6. Duda R., Hart P. Raspoznavanie obrazov i analiz scen: Per s angl. V.L. Stefanyuka. M.: Mir. 1976. S. 511 (in Russian).
7. Kovalevskij V.A. Metody optimal'nyh reshenij v raspoznavanii izobrazhenij. M.: Nauka. 1976. S. 328 (in Russian).
8. Tu Dzh., Gonsales R. Principy raspoznavaniya obrazov: Per. s angl. M. Zhuravleva. M.: Mir. 1978. S. 412 (in Russian).
9. Selekciya i raspoznavanie na osnove lokacionnoj informacii. Pod red. A.L. Gorelik. M. Radio i svyaz'. 1990. S. 236 (in Russian).
10. Gorelik A.L., Skripkin V.A. Metody raspoznavaniya. M.: Vysshaya shkola. 1984. S. 208 (in Russian).
11. Dostovalov M.Yu., Lifanov A.S., Musinyanc T.G. Sistema avtomatizirovannogo raspoznavaniya ob"ektov na osnove issledovaniya radiolokacionnyh portretov po izobrazheniyam RSA KOMPAKT-100. Issledovaniya Zemli iz kosmosa. 2005. № 1 (in Russian).
12. Diemunsch J., Wissinger J. MSTAR model-based automatic target recognition: search technology for a robust ATR. Algorithms for Synthetic Aperture Radar Imagery V.E. Zeinio, Eds. Proceedings of SPIE. 1998. V. 3370.
13. Egorova E.V., Rybakov A.N., Aksyaitov M.H. Optimal'naya obrabotka informacii pri raspo-znavanii i identifikacii radiolokacionnyh celej: Monografiya. pod red. S.G. Tolmacheva D.V. Klopova; Minobrnauki Rossii, RTU MIREA. Tambov: Konsaltingovaya kompaniya «Yukom». 2021. S. 136 (in Russian).
14. Aksyaitov M.H., Egorova E.V., Martynov N.V., Rybakov A.N. Obnaruzhenie malokontrastnyh celej. Uspekhi sovremennoj radioelektroniki 2017. № 1. S. 23–26 (in Russian).
15. Dorosinskij L.G., Lysenko T.M. Analiz adaptivnogo algoritma obnaruzheniya konturov na ma-lokontrastnyh radiolokacionnyh izobrazheniyah. Issledovanie zemli iz kosmosa. 1988. № 2. S. 99–103 (in Russian).
16. Dorosinskij L.G. Vybor metoda sovmestnogo izmereniya uglovyh koordinat neskol'kih is-tochnikov izlucheniya. Radiotekhnika. 1987. № 11. S. 43–45 (in Russian).