Makine Öğrenimi Tabanlı SAR ATR'de ISAR Eğitim Verilerinin Kullanımına İlişkin Bir Fizibilite Analizi

Year 2023, Volume: 15 Issue: 3, 302 - 308, 31.12.2023

Enes Yiğit , Şevket Demirci , Umut Özkaya

https://doi.org/10.29137/umagd.1402020

Abstract

Otomatik hedef tanıma (ATR) için sentetik açıklıklı radar (SAR) görüntülerinin işlenmesi, özellikle askeri gözetlemede kritik bir uygulamadır. Özellikle, bu görev için çok sayıda makine öğrenimine dayalı SAR ATR yöntemi önerilmiştir. Ancak tüm bu yöntemlerin veri eğitimi ve test aşamaları, araştırılan hedefin SAR imzalarının kullanılmasına dayanmaktadır. Radar hedeflerinin yüksek değişkenliği göz önüne alındığında, bu tür imza verilerinin elde edilmesinin maliyetli ve zaman alıcı bir süreç olduğu açıktır. Bu çalışmada, bu nedenle, SAR ATR'de ters SAR (ISAR) eğitim verilerinin kullanımının ilk kez bir fizibilite analizi yapılmıştır. Eğitimlerde üç farklı aracın döner tabla ISAR ve dairesel SAR görüntüleri kullanılmakta ve kamuya açık MSTAR veri setinde benzer üç hedefin SAR görüntüleri aracılığıyla testler gerçekleştirilmektedir. Ayrıca, KNN, SVM ve ANN olmak üzere en önde gelen üç makine öğrenme yöntemi, GLRLM, GLSZM ve GLCM olmak üzere üç farklı özellik çıkarma algoritması ile birlikte kullanılmaktadır. Elde edilen sonuçlar, GLCM+SVM algoritma çiftinin %85 doğrulukla en etkili model olduğunu ortaya koymaktadır.

Keywords

Otomatik Hedef Tanıma, Sentetik Açıklıklı Radar, Ters Sentetik Açıklıklı Radar, Yapay Zeka

A Feasibility Analysis of the Use of ISAR Training Data in Machine Learning-Based SAR ATR

Abstract

Processing of synthetic aperture radar (SAR) images for automatic target recognition (ATR) is a critical application especially in military surveillance. In particular, numerous machine learning-based SAR ATR methods have been proposed for this task. However, data training and testing stages of all these methods are based on the exploitation of SAR signatures of the target under investigation. Considering the high variability of radar targets, obtaining such signature data is obviously a costly and time consuming process. In this study, therefore, a feasibility analysis of the use of inverse-SAR (ISAR) training data in SAR ATR has been made for the first time. The turntable ISAR and circular SAR images of three different vehicles are used in training and testing is performed by means of SAR images of three similar targets within the publicly available MSTAR dataset. Also, three most prominent machine learning methods, namely KNN, SVM and ANN are used in conjunction with three different feature extraction algorithms namely, GLRLM, GLSZM and GLCM. The obtained results reveal that the GLCM+SVM algorithm pair is the most effective model with 85% accuracy.

Keywords

Automatic Target Recognition, Synthetic Aperture Radar, Inverse Synthetic Aperture Radar, Artificial Intelligence

References

• Blacknell, D. & Vignaud, L. (2013). EN-SET-172-01 ATR of Ground Targets: Fundamentals and Key Challenges. Radar Autom. Target Recognit. Non-Cooperative Target Recognit., 1–36. [Online]. Available: https://www.sto.nato.int/publications/STO Educational Notes/RTO-EN-SET-172/EN-SET-172-01.pdf.
• Copsey, K. D. (2004). Bayesian Approaches for Robust Automatic Target Recognition. Dr. Philos. Univ. London.
• Cui, X. C., Fu, Y. W., Su, Y., & Chen, S. W. (2023). Physical Parameters Joint Estimation of Satellite Parabolic Antenna with Key Frame Pol-ISAR Images. IEEE Transactions on Geoscience and Remote Sensing. doi: 10.1109/TGRS.2023.3335960.
• Demirci, S., Yigit, E. & Ozdemir, C. (20115). Wide-field circular SAR imaging: An empirical assessment of layover effects. Microw. Opt. Technol. Lett., 57(2), 489–497. doi: 10.1002/mop.28884.
• Demirci, S., & Izumi, Y. (2023). Application of H/ᾱ Decomposition to Limited and Dual-Polarimetric 3D SAR Data of Civilian Vehicles. IEEE Access. doi: 10.1109/ACCESS.2023.3301619
• Dong, G., Kuang, G. Wang, N., Zhao, L. & Lu, J. (2015). SAR Target Recognition via Joint Sparse Representation of Monogenic Signal, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 8(7), 3316–3328. doi: 10.1109/JSTARS.2015.2436694.
• Dong, G., Kuang, G., Wang, N. & Wang, W. (2017). Classification via sparse representation of steerable wavelet frames on grassmann manifold: Application to target recognition in SAR image. IEEE Trans. Image Process., 26(6), 2892–2904. doi: 10.1109/TIP.2017.2692524.
• Duysak, H., Özkaya, U. & Yigit, E. (2020). Makine Öğrenmesi Yöntemleri ile Tahıl Yüzey Sınıflaması. Eur. J. Sci. Technol. doi: 10.31590/ejosat.802719.
• Duysak, H. & Yigit, E. (2020). Machine learning based quantity measurement method for grain silos, Meas. J. Int. Meas. Confed., 152. doi: 10.1016/j.measurement.2019.107279.
• Gotcha Volumetric SAR Data Set Overview. https://www.sdms.afrl.af.mil/index.php?collection=gotcha (accessed Sep. 08, 2021).
• Liu, H. & Li, S. (2013). Decision fusion of sparse representation and support vector machine for SAR image target recognition. Neurocomputing, 113, 97–104. doi: 10.1016/J.NEUCOM.2013.01.033.
• Martone, A. Innocenti, R. & Ranney, K. (2009). Moving Target Indication for Transparent Urban Structures, US Army Res. Lab. Adelphi United States.
• Miao, X. & Liu, Y. (2021). Target Recognition of SAR Images Based on Complex Bidimensional Empirical Mode Decomposition. Sci. Program. doi: 10.1155/2021/6642316.
• MSTAR Public Targets. https://www.sdms.afrl.af.mil/index.php?collection=mstar&page=targets (accessed Sep. 03, 2021).
• Novak, L., Owirka, G. Netishen, M. (1993). Performance of a High-Resolution Polarimetric SAR Automatic Target Recognition System, Lincoln Lab. J., 6(1).
• Novak, L. M., Owirka, G. J. & Brower, W. S. (1998). Efficient multi-target SAR ATR algorithm. Conf. Rec. Asilomar Conf. Signals, Syst. Comput., 1, 3–13. doi: 10.1109/ACSSC.1998.750815.
• Song, J., Kim, D. J., Hwang, J. H., Kim, H., Li, C., Han, S., & Kim, J. (2023). Effective Vessel Recognition in High Resolution SAR Images Utilizing Quantitative and Qualitative Training Data Enhancement from Target Velocity Phase Refocusing. IEEE Transactions on Geoscience and Remote Sensing. doi: 10.1109/TGRS.2023.3346171
• Özkaya, U. (2020). Automatic Target Recognition (ATR) from SAR Imaginary by Using Machine Learning Techniques. Avrupa Bilim ve Teknol. Dergisi, 165–169. doi: 10.31590/EJOSAT.802811.
• Özkaya, U., Yigit, E., Seyfi, L., Özturk, S. & Singh, D. (2021). Comparative regression analysis for estimating resonant frequency of C-like patch antennas, Math. Probl. Eng., doi: 10.1155/2021/6903925.
• U.S. Air Force, Sesnsor Data Management System (SDMS), (1997), GTRI dataset.” https://www.sdms.afrl.af.mil/index.php?collection=gtri (accessed Sep. 08, 2021).
• Vertiy, A., Cetinkaya, H., Panin, S., Pavlyuchenko, A., Tekbas, M., Unal, A., Kizilhan, A., Kaya, A., Ozdemir, C., Demirci, S. & Yigit, E. (2011). Image reconstruction in SAR, ISAR and tomography applications at millimeter-wave band. Microwaves, Radar and Remote Sensing Symposium, doi: 10.1109/MRRS.2011.6053631.
• Yigit, E. (2018). Operating Frequency Estimation of Slot Antenna by Using Adapted kNN Algorithm, Int. J. Intell. Syst. Appl. Eng., 6(1), 29–32, doi: 10.1039/b000000x.
• Yigit, E. (2020). A translational motion compensation technique for inverse synthetic aperture radar images using multi-objective particle swarm optimization algorithm. Microw. Opt. Technol. Lett., 62(6), doi: 10.1002/mop.32314.