Impact of the Radar Image Resolution of Military Objects on the Accuracy of their Classification by a Deep Convolutional Neural Network
https://doi.org/10.32603/1993-8985-2022-25-1-36-46
Abstract
Introduction. Deep convolutional neural networks are considered as one of the most promising tools for classifying small-sized objects on radar images. However, no systemic study has been reported so far on the dependence between the classification accuracy achieved by convolutional neural networks and such an important image characteristic as resolution.
Aim. Determination of a dependence between of the accuracy of classifying military objects by a deep convolutional neural network and the resolution of their radar images.
Materials and methods. An eight-layer convolutional neural network was designed, trained and tested using the Keras library and Tensorflow 2.0 framework. For training and testing, the open part of the standard MSTAR dataset comprising ten classes of military objects radar images was used. The initial weight values of the MobileNetV1 and Xception networks used for a comparative assessment of the achieved classification accuracy were obtained from the training results on the Imagenet.
Results. The accuracy of classifying military objects decreases rapidly along with a deterioration in resolution, amounting to 97.91, 90.22, 79.13, 52.2 and 23.68 % at a resolution of 0.3, 0.6, 0.9, 1.5 and 3 m, respectively. It is shown that the use of pretrained MobileNetV1 and Xception networks does not lead to an improvement in the classification accuracy compared to a simple VGG-type network.
Conclusion. Effective recognition of military objects at a resolution worse than one meter is practically impossible. The classification accuracy of deep neural networks depends significantly on the difference in the image resolution of the training and test sets. A significant increase in the resistance of the classification accuracy to changes in the resolution can be achieved by training on a set of images with different resolutions.
About the Author
I. F. Kupryashkin
Military Educational and Scientific Center of the Air Force "N. E. Zhukovsky and Y. A. Gagarin Air Force Academy"
Russian Federation
Russian Federation
Ivan F. Kupryashkin – Dr Sci. (Eng.) (2017), Assosiate Professor (2011), Head of the Departament of Сombat Use of Electronic Warfare Systems (with Aerospace Control Systems and Guided Weapons)
54 A, Starykh Bolshevikov Str., Voronezh 394064
References
1. Zhu X., Montazeri S., Ali M., Hua Yu., Wang Yu., Mou L., Shi Yi., Xu F., Bamler R. Deep Learning Meets SAR. URL: https://arxiv.org/pdf/2006.10027.pdf (accessed 20.12.2021)
2. Wang H. H., Chen S., Xu F., Jin Y.-Q. Application of Deep-Learning Algorithms to MSTAR. IEEE Intern. Geo-science and Remote Sensing Symp. (IGARSS), Milan, Italy, 26–31 July 2015. IEEE, 2015, pp. 3743–3745. doi: 10.1109/IGARSS.2015.7326637
3. Chen S., Wang H., Xu F., Jin Y.-Q. Target Classifica-tion Using the Deep Convolutional Networks for SAR Images. IEEE Transaction Geoscience and Remote Sens-ing. 2016, vol. 54, no. 8, pp. 4806–4817. doi: 10.1109/ TGRS.2016.2551720
4. Anas H., Majdoulayne H., Chaimae A., Nabil S. M. Deep Learning for SAR Image. Intelligent Systems and Applications. Springer, Cham., 2020, pp. 890–898. doi: 10.1007/978-3-030-29516-5_67
5. Chen S., Wang H. SAR Target Recognition Based on Deep Learning // Intern. Conf. on Data Science and Advanced Analytics (DSAA). Shanghai, China, 30 Oct. – 1 Nov. 2014. IEEE, 2014, pp. 541–547. doi: 10.1109/DSAA.2014.7058124
6. Coman C., Thaens R. A Deep Learning SAR Target Classification Experiment on MSTAR Dataset. 19th Intern. Radar Symp. (IRS). Bonn, Germany, 20–22 June 2018. IEEE, 2018. P. 1–6. doi: 10.23919/IRS.2018.8448048
7. Furukawa H. Deep Learning for End-to-End Automat-ic Target Recognition from Synthetic Aperture Radar Image-ry. Available at: https://arxiv.org/pdf/1801.08558.pdf (ac-cessed 20.12.2021)
8. Profeta A., Rodriguez A., Clouse H. S. Convolutional Neural Networks for Synthetic Aperture Radar Classifica-tion. Proc. SPIE 9843, Algorithms for Synthetic Aperture Radar Imagery XXIII. 2016, 98430M. doi: 10.1117/12.2225934
9. Wang Z., Xu X. Efficient Deep Convolutional Neural Networks Using CReLU for ATR with Limited SAR Images. The J. of Engineering. 2019, vol. 2019, no. 21, pp. 7615–7618. doi: 10.1049/joe.2019.0567
10. Wilmanski M., Kreucher C., Lauer J. Modern Ap-proaches in Deep Learning for SAR ATR. Proc. SPIE 9843. Algorithms for Synthetic Aperture Radar Imagery XXIII. 2016, 98430N. doi: 10.1117/12.2220290
11. Xie Yi., Dai W., Hu Z., Liu Yi., Li C., Pu X. A Novel Convolutional Neural Network Architecture for SAR Tar-get Recognition. J. of Sensors. 2019, art. 1246548. doi: 10.1155/2019/1246548
12. Xinyan F., Weigang Z. Research on SAR Image Target Recognition Based on Convolutional Neural Net-work. J. of Physics: Conf. Series. 2019, ser. 1213, 042019. doi: 10.1088/1742-6596/1213/4/042019
13. Zhai J., Dong G., Chen F., Xie X., Qi C., Li L. A Deep Learning Fusion Recognition Method Based On SAR Im-age Data. Procedia Computer Science. 2019, vol. 147, pp. 533–541. doi: 10.1016/j.procs.2019.01.229
14. Gao F., Huang T., Sun J., Wang J., Hussain A., Yang E. A New Algorithm of SAR Image Target Recognition based on Improved Deep Convolutional Neural Network. Cognitive Computation. 2019, vol. 11, pp. 809–824. doi: 10.1007/s12559-018-9563-z
15. Malmgren-Hansen D., Engholm R., Østergaard Pedersen M. Training Convolutional Neural Net-works for Translational Invariance on SAR ATR. Proc. of EUSAR 2016: 11th European Conf. on Synthetic Aperture Radar, Ham-burg, Germany, 6–9 Jun 2016. IEEE, 2016, pp. 459–462.
16. Borodinov A. A., Myasnikov V. V. Comparison of Radar Image Classification Algorithms for Various Pre-processing Methods Based on MSTAR Data. Proc. of the IV Intern. Conf. and Youth School "Information Tech-nology and Nanotechnology" (ITNT-2018). Samara, New Equipment, 2018, pp. 586–594. (In Russ.)
17. Kechagias-Stamatis O., Aouf N. Automatic Target Recognition on Synthetic Aperture Radar Imagery: A Survey. URL: https://arxiv.org/ftp/arxiv/papers/2007/2007.02106.pdf (accessed 20.12.2021)
18. Chollet F. Deep Learning with Python. NY, Man-ning, 2017, 384 p.
19. Nikolenko S., Kadurin A., Arkhangel’skaya E. Glubokoe obuchenie [Deep Learning]. SPb., Piter, 2018, 480 p. (In Russ.).
20. Simonyan K., Zisserman A. Very Deep Convolutional Networks For Large-Scale Image Recognition. Available at: https://arxiv.org/pdf/1409.1556.pdf (accessed 20.12.2021)
21. Huang Z., Pan Z., Lei B. What, Where and How to Transfer in SAR Target Recognition Based on Deep CNNs. Available at: https://arxiv.org/pdf/1906.01379.pdf (accessed 20.12.2021)
22. Chollet F. Xception: Deep Learning with Depth-wise Separable Convolutions. Available at: https://arxiv.org/pdf/1610.02357.pdf (accessed 20.12.2021).
23. Howard A.G., Zhu M., Chen B., Kalenichenko D. MobileNets: Efficient Convolutional Neural Net-works for Mobile Vision Applications. Available at: https://arxiv.org/pdf/1704.04861.pdf (accessed 20.12.2021)
Review
For citations:
Kupryashkin I.F. Impact of the Radar Image Resolution of Military Objects on the Accuracy of their Classification by a Deep Convolutional Neural Network. Journal of the Russian Universities. Radioelectronics. 2022;25(1):36-46. (In Russ.) https://doi.org/10.32603/1993-8985-2022-25-1-36-46
Views:
506
Email this article 