Cybernetics and Computer Engineering, 2022, 3(209)
1International Research and Training Center for Information Technologies and Systems of the National Academy of Sciences of Ukraine and Ministry of Education and Science of Ukraine, 40, Acad. Glushkova av., Kyiv, 03187, Ukraine
2Taras Shevchenko National University of Kyiv,
Faculty of Computer Sciences and Cybernetics,
4d, Acad. Glushkova av., Kyiv, 03022, Ukraine
MEANS FOR A CLASSIFICATION TECHNOLOGY OF SYNTHETIC RADAR IMAGES OF OBJECTS HAVING COMPLEX SHAPES
Introduction. Currently, research into the synthesis of wave images of reflected sound and radio signals has been actively carried out, due to the fact a successful attempt to determine the type of an object for which there is such an image requires either a very large sample base or an intelligent recognition tool. An attempt is made to analyze and recognize the type of an object of a complex shape (using ships as example) with the aim of its further use in applied tasks such as creation of homing heads for anti-ship missiles.
The purpose of the paper is to simplify and speed up the process of classifying objects having complex shapes based on their reflected radar images. For this purpose, we consider synthesized images generated on the basis of facet models. Then, on the basis of synthesized images, recognition is performed using neural networks.
Results. It is shown that the method developed for recognition of synthesized images has high reliability, and allows for building of a technology in the future. The elaborated model of image generation provides for a possibility of conducting experiments exclusively in a digital form, making thereby expensive live experiments unnecessary.
Conclusions. Despite very good results from a mathematical point of view, and in spite of the available convenient tools, such as faceted models for creating radar images, the task still requires further research, since the final product (technology) must be applied in the area where the cost of an error is very high. As for now, the development of the neural network approach looks the most promising.
Keywords: facet model; remote sensing; underlying surface; radar image
1. Volkov O.Ye.,Taranukha V.Yu, Linder Ya.M., Komar M.M., Volosheniuk D.O. Devising an acoustic method for investigation of a complex form object parameters. Cyb. and Comp. Eng. 2021, No 4 (206), pp. 39-53.
2. Voskresensky D.I. (ed.) Microwave devices and antennas. Design of phased antenna arrays Textbook for universities. – 4th edition, revised and enlarged. – Moscow: Radiotekhnika, 2012. – 744 p. (in Russian)
3. Kalinin T.V., Bartsevich A.B., Petrov S.A., Khrestinin D.V. Software package for modeling radar recognition systems, Software Products and Systems. 2017. V. 30, no. 4. pp. 733-738 (in Russian).
4. Khrichov V.S., Legenkiy M.M. A faceted model of an object of a folding mold for the opening of an electromagnetic rosette. Bulletin of Kharkiv National University named after V. N. Karazin. Series “Radiophysics and electronics”, (2019). V. 28, pp. 44-52. (in Russian)
5. Chen V.C., Hao Ling. Time-Frequency transforms for radar imaging and signal analysis. Artech House, 2002, 234 p. ISBN 978-1580532884
6. Rihaczek A.W., Hershkowitz S.J. Theory and practice of radar target identification. Artech House, 2000, 738 p. ISBN 978-1580530811
7. Konovalyuk M.A., Kuznetsov Yu.V., Baev A.B. Identification of parameters of a multi-point target from a complex radar image in broadband radar. 3rd International Conference “Acousto-optical and radar methods of measurement and information processing” (ARMIMP- 2009), Suzdal, September 22-24, 2009, p. 48-52. (in Russian)
8. Munoz-Ferreras J.M., Perez-Martinez F. On the Doppler Spreading Effect for the Range-Instantaneous-Doppler Technique in Inverse Synthetic Aperture Radar Imagery. IEEE Geoscience and Remote Sensing Letters. Vol. 7, No. 1, pp. 180-184, Jan. 2010.
9. Graves A., Mohamed A., Hinton G.. Speech Recognition with Deep Recurrent Neural Networks. ICASSP. IEEE Int. Conf. Acoust. Speech Signal Process. – Proc. Vol. 38, 2013, pp. 6645-6649.
10. Malhotra P., Vig L., Shroff G., Agarwal P. Long Short Term Memory Networks for Anomaly Detection in Time Series. European Symposium on Artificial Neural Networks. Computational Intelligence and Machine Learning. Bruges (Belgium), 22-24 April 2015, pp. 89-94
11. Siami-Namini S., Siami Namin A. Forecasting Economics and Financial Time Series: ARIMA vs. LSTM. 2018.
12. Hochreiter S., Schmidhuber J. Long Short-term Memory. Neural Comput., 1997,V. 9, pp. 1735-1780.
13. Schuster M., Paliwal K.. Bidirectional recurrent neural networks. Signal Process. IEEE Trans. vol. 45, Dec. 1997, pp. 2673-2681.
14. Cui Z., Ke R., Pu Z., Wang Y. Deep Bidirectional and Unidirectional LSTM Recurrent Neural Network. Network-wide Traffic Speed Prediction. 2018.
15. Bahdanau D., Cho K., Bengio Y. Neural Machine Translation by Jointly Learning to Align and Translate. Sep. 2014.
16. Bracewell R. The Fourier Transform and Its Applications, 2nd ed. McGraw-Hill. Jan. 1978, 444 p.
17. Kingma D.P, Ba J. Adam: A Method for Stochastic Optimization. https://arxiv.org/abs/1412.6980..
HOW TO CITE:
Volkov O.Ye. , Bogachuk Yu.P., Linder Ya.M., Taranukha V.Yu., Voloshenyuk D.O. Means for a Classification Technology of Synthetic Radar Images of Objects Having Complex Shapes. Cybernetics and Computer Engineering, 2022, no 3(209), pp.5-21.