Zhurnal Radioelektroniki - Journal of Radio Electronics. eISSN 1684-1719. 2021. ¹11
Contents

Full text in Russian (pdf)

Russian page

 

DOI: https://doi.org/10.30898/1684-1719.2021.11.14

UDC: 621.396.96

 

COMPARATIVE RESULTS OF THE CLASSIFICATION ACCURACY OF MSTAR DATASET RADAR IMAGES BY CONVOLUTIONAL NEURAL NETWORKS WITH DIFFERENT ARCHITECTURES

 

I. F. Kupryashkin

 

Military Educational and Scientific Center of the Air Force «N.E. Zhukovsky and Y.A. Gagarin Air Force Academy»

394064, Voronezh, Starykh Bolshevikov St., 64

 

The paper was received October 20, 2021.

 

Abstract. The results of MSTAR objects ten-classes classification using a VGG-type deep convolutional neural network with eight convolutional layers are presented. The maximum accuracy achieved by the network was 97.91%. In addition, the results of the MobileNetV1, Xception, InceptionV3, ResNet50, InceptionResNetV2, DenseNet121 networks, prepared using the transfer learning technique, are presented. It is shown that in the problem under consideration, the use of the listed pretrained convolutional networks did not improve the classification accuracy, which ranged from 93.79% to 97.36%. It has been established that even visually unobservable local features of the terrain background near each type of object are capable of providing a classification accuracy of about 51% (and not the expected 10% for a ten-alternative classification) even in the absence of object and their shadows. The procedure for preparing training data is described, which ensures the elimination of the influence of the terrain background on the result of neural network classification

Key words: deep convolutional neural network, transfer learning, radar image, classification accuracy

References

1. Alzubaidi L., Zhang J., Humaidi A.J., Al-Dujaili A., Duan Y., Al-Shamma O., Santamaria J., Fadhel M.A., Al-Amidie M., Farhan L. Review Of Deep Learning: Concepts, CNN Architectures, Challenges, Applications, Future Directions. Journal of Big Data. 2021. V.8. ¹53.

2. Rawat W., Wang Z. Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review. Neural Computation. 2017. V.29. P.2352-2449. https://doi.org/10.1162/neco_a_00990

3. Sozykin A.V. An overview of methods for deep learning in neural networks. Vestnik YUUrGU. Seriya «Vychislitel'naya matematika i informatika» [Bulletin of the South Ural State University. Series «Computational Mathematics and Computer Science»]. 2017. V.6. ¹3. P.28-59. https://doi.org/10.14529/cmse170303 (In Russian)

4. Zhu X., Montazeri S., Ali M., Hua Yu., Wang Yu., Mou L., Shi Yi., Xu F., Bamler R. Deep Learning Meets SAR. ArXiv. https://arxiv.org/abs/2006.10027

5. Wang P., Patel V.M. Generating High Quality Visible Images from SAR Images Using CNNs. ArXiv. 2018. https://arxiv.org/abs/1802.10036

6. Rittenbach A., Walters J.P. RDAnet: A Deep Learning Based Approach for Synthetic Aperture Radar Image Formation. ArXiv. 2020. https://arxiv.org/abs/2001.08202

7. Hu C., Wang L., Li Z., Zhu D. Inverse Synthetic Aperture Radar Imaging Using a Fully Convolutional Neural Network. IEEE Geoscience and Remote Sensing Letters. 2020. V.17. ¹7. P.1203-1207. https://doi.org/10.1109/LGRS.2019.2943069

8. Wang H., Chen S., Xu F., Jin Y.-Q. Application of Deep-Learning Algorithms to MSTAR Data. 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS). 2015. P.3743-3745. https://doi.org/10.1109/IGARSS.2015.7326637

9. Chen S., Wang H., Xu F., Jin Y.-Q. Target Classification Using the Deep Convolutional Networks for SAR Images. IEEE Transaction Geoscience and Remote Sensing. 2016. V.54. ¹8. P.4806-4817. https://doi.org/10.1109/TGRS.2016.2551720

10. Anas H., Majdoulayne H., Chaimae A., Nabil S.M. Deep Learning for SAR Image Classification. Intelligent Systems and Applications. 2020. P.890-898. https://doi.org/10.1007/978-3-030-29516-5_67

11. Chen S., Wang H. SAR Target Recognition Based on Deep Learning. 2014 International Conference on Data Science and Advanced Analytics (DSAA). 2014. P.541-547. https://doi.org/10.1109/DSAA.2014.7058124

12. Coman C., Thaens R. A Deep Learning SAR Target Classification Experiment on MSTAR Dataset. 2018 19th International Radar Symposium (IRS). 2018. P.1-6. https://doi.org/10.23919/IRS.2018.8448048

13. Furukawa H. Deep Learning for End-to-End Automatic Target Recognition from Synthetic Aperture Radar Imagery. ArXiv. https://arxiv.org/abs/1801.08558

14. Profeta A., Rodriguez A., Clouse H.S. Convolutional Neural Networks for Synthetic Aperture Radar Classification. Proc. SPIE 9843, Algorithms for Synthetic Aperture Radar Imagery XXIII, 98430M. https://doi.org/10.1117/12.2225934

15. Wang Z., Xu X. Efficient deep convolutional neural networks using CReLU for ATR with limited SAR images. The Journal of Engineering. 2019. V.2019. ¹21. P.7615-7618. https://doi.org/10.1049/joe.2019.0567

16. Wilmanski M., Kreucher C., Lauer J. Modern Approaches in Deep Learning for SAR ATR. Proc. SPIE 9843, Algorithms for Synthetic Aperture Radar Imagery XXIII, 98430N. https://doi.org/10.1117/12.2220290

17. Xie Yi., Dai W., Hu Z., Liu Yi., Li C., Pu X. A Novel Convolutional Neural Network Architecture for SAR Target Recognition. Journal of Sensors. 2019. https://doi.org/10.1155/2019/1246548

18. Xinyan F., Weigang Z. Research on SAR Image Target Recognition Based on Convolutional Neural Network. Journal of Physics: Conference Series. 2019. V.1213(2019) 042019. https://doi.org/10.1088/1742-6596/1213/4/042019

19. Zhai J., Dong G., Chen F., Xie X., Qi C., Li L. A Deep Learning Fusion Recognition Method Based On SAR Image Data. 2018 International Conference on Identifiation, Information and Knowledge in the Internet of Things. Procedia Computer Science. 2019. V.147. P.533-541. https://doi.org/10.1016/j.procs.2019.01.229

20. LeCun Y., Bottou L., Bengio Y., Haffner P. Gradient-Based Learning Applied to Document Recognition. Proceedings Of The IEEE. 1998. V.86. ¹11. P.2278-2324. https://doi.org/10.1109/5.726791

21. Krizhevsky A., Sutskever I., Hinton G.E. ImageNet Classification with Deep Convolutional Neural Networks. Communications of the ACM. 2017. V.60. ¹6. P.84-90. https://doi.org/10.1145/3065386

22. 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. V.11. P.809-824. https://doi.org/10.1007/s12559-018-9563-z

23. Malmgren-Hansen D., Engholm R., Østergaard Pedersen M. Training Convolutional Neural Networks for Translational Invariance on SAR ATR.  Proceedings of EUSAR 2016: 11th European Conference on Synthetic Aperture Radar. 2016. P.459-462.

24. Borodinov A.A., Myasnikov V.V. Comparison of classification algorithms for radar images with different preprocessing methods using the example of the MSTAR base. Proceedings of the IV International Conference and Youth School «Information Technology and Nanotechnology» (ITNT-2018). Samara, New Equipment. 2018. P.586-594. (In Russian)

25. Tatuzov A.L. Nejronnye seti v zadachah radiolokacii. Kn. 28. [Neural Networks in Radars. B. 28.]. Moscow, Radiotekhnika Publ. 2009. 432 p. (In Russian)

26. Bilenko S.V., Cheredeyev K.Yu., Zograbyan M.K. Prospects for the use of deep neural networks in radiolocation. Voprosy radioelektroniki [Issues of Radioelectronics]. 2017. ¹1. P.57-63 (In Russian)

27. Kazachkov E.A., Matyugin S.N., Popov I.V., Sharonov V.V. Detection and classification of small-scale objects in images obtained by synthetic-aperture radar stations. Vestnik Koncerna VKO «Almaz - Antej» [Journal of «Almaz - Antey» Air and Space Defence Corporation]. 2018. ¹1. P.93-99 (In Russian).

28. Chorbaa N.A., Le Anh Tu, Tolstoy I.M. Comparative Analysis of Methods for Detecting Objects on Radar Images Using Neural Networks. Nauchnyj rezul'tat. Informacionnye tekhnologii [Research Result. Information Technologies]. 2020. V.5. ¹4. P.15-25. https://doi.org/10.18413/2518-1092-2020-5-4-0-3 (In Russian)

29. Kechagias-Stamatis O., Aouf N. Automatic Target Recognition on Synthetic Aperture Radar Imagery: A Survey. ArXiv. https://arxiv.org/abs/2007.02106

30. Chollet F. Glubokoe obuchenie na Python [Deap Learning with Python]. Saint-Petersburg, Piter Publ. 2018. 400 p. (In Russian).

31. Nikolenko S., Kadurin A., Arkhangel’skaya E. Glubokoe obuchenie [Deep Learning]. Saint-Petersburg, Piter Publ. 2018. 480 p. (In Russian).

32. Agarwal T., Sugavanam N., Ertin E. Sparse Signal Models for Data Augmentation in Deep Learning ATR. ArXiv. https://arxiv.org/abs/2012.09284

33. Furukawa H. Deep Learning for Target Classification from SAR Imagery Data Augmentation and Translation Invariance. ArXiv. https://arxiv.org/abs/1708.07920

34. Malmgren-Hansen D., Kusk A., Dall J., Nielsen A. A., Engholm R., Skriver H. Improving SAR Automatic Target Recognition Models with Transfer Learning from Simulated Data. IEEE Geoscience and Remote Sensing Letters. 2017. V.14(9). P.1484-1488. https://doi.org/10.1109/LGRS.2017.2717486

35. Wang J., Virtue P., Yu S.X. Joint Embedding and Classification for SAR Target Recognition. ArXiv. https://arxiv.org/abs/1712.01511

36. Simonyan K., Zisserman A. Very Deep Convolutional Networks For Large-Scale Image Recognition. ArXiv. https://arxiv.org/abs/1409.1556

37. Huang Z., Pan Z., Lei B. What, Where and How to Transfer in SAR Target Recognition Based on Deep CNNs. ArXiv. https://arxiv.org/abs/1906.01379

38. Szegedy C., Liu W., Jia Ya., Sermanet P., Reed S., Anguelov D., Erhan D., Vanhoucke V., Rabinovic A. Going deeper with convolutions. ArXiv. https://arxiv.org/abs/1409.4842

39. Szegedy C., Vanhoucke V., Ioffe S., Shlens J., Wojna Z. Rethinking the Inception Architecture for Computer Vision. ArXiv. https://arxiv.org/abs/1512.00567

40. He K., Zhang X., Ren S., Sun J. Deep Residual Learning for Image Recognition. ArXiv. https://arxiv.org/abs/1512.03385

41. Szegedy C., Vanhoucke V., Ioffe S. Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning. ArXiv. https://arxiv.org/abs/1602.07261

42. Chollet F. Xception: Deep Learning with Depthwise Separable Convolutions. ArXiv. https://arxiv.org/abs/1610.02357

43. Huang G., Liu Z., van der Maaten L. Densely Connected Convolutional Networks. ArXiv. https://arxiv.org/abs/1608.06993

44. Howard A.G., Zhu M., Chen B., Kalenichenko D. MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. ArXiv. https://arxiv.org/abs/1704.04861

For citation:

Kupryashkin I.F. Comparative results of the classification accuracy of MSTAR dataset radar images by convolutional neural networks with different architectures. Zhurnal Radioelektroniki [Journal of Radio Electronics] [online]. 2021. ¹11. https://doi.org/10.30898/1684-1719.2021.11.14