Journal of the Korea Institute of Military Science and Technology (한국군사과학기술학회지)

The Korea Institute of Military Science and Technology (한국군사과학기술학회)
권호 표지
DOI QR코드

DOI QR Code

Performance Analysis of Deep Learning-Based Detection/Classification for SAR Ground Targets with the Synthetic Dataset

합성 데이터를 이용한 SAR 지상표적의 딥러닝 탐지/분류 성능분석

  • Ji-Hoon Park (Advanced Defense Science & Technology Research Institute, Agency for Defense Development)
  • 박지훈 (국방과학연구소 국방첨단과학기술연구원)
  • Received : 2023.08.16
  • Accepted : 2024.01.16
  • Published : 2024.04.05
Download PDF
⟨ Previous Next ⟩

Abstract

Based on the recently developed deep learning technology, many studies have been conducted on deep learning networks that simultaneously detect and classify targets of interest in synthetic aperture radar(SAR) images. Although numerous research results have been derived mainly with the open SAR ship datasets, there is a lack of work carried out on the deep learning network aimed at detecting and classifying SAR ground targets and trained with the synthetic dataset generated from electromagnetic scattering simulations. In this respect, this paper presents the deep learning network trained with the synthetic dataset and applies it to detecting and classifying real SAR ground targets. With experiment results, this paper also analyzes the network performance according to the composition ratio between the real measured data and the synthetic data involved in network training. Finally, the summary and limitations are discussed to give information on the future research direction.

Keywords

Acknowledgement

이 논문은 2023년 정부의 재원으로 수행된 연구 결과임.

References

  1. MSTAR Overview, https://www.sdms.afrl.af.mil/index.php?collection=mstar, accessed (2023)
  2. S. Chen, et. al., "Target Classification Using the Deep Convolutional Networks for SAR Images," IEEE Trans. Geoscience and Remote Sensing, Vol. 54, No. 8, pp. 4806-4817, 2016. https://doi.org/10.1109/TGRS.2016.2551720
  3. R. Xue, et. al., "Spatial-Temporal Ensemble Convolution for Sequence SAR Target Classification," IEEE Trans. Geoscience and Remote Sensing, Vol. 59, No. 2, pp. 1250-1262, 2021. https://doi.org/10.1109/TGRS.2020.2997288
  4. S. Wang, et. al., "YOLO-SD : Small Detection in SAR Images by Multi-Scale Convolution and Feature Transformer Module," Remote Sensing, Vol. 14, No. 20, 2022.
  5. Y. Gao, et. al., "Improved YOLO v4 Based on Attention Mechanism for Ship Detection for Ship Detection in SAR Images," IEEE Access, Vol. 10, pp. 23785-23797, 2022. https://doi.org/10.1109/ACCESS.2022.3154474
  6. N. Inkawhich, et. al., "Bridging a Gap in SARATR : Training on Fully Synthetic and Testing on Measured Data," IEEE J. Selected Topics in Applied Earth Observations and Remote Sensing, Vol. 14, pp. 2942-2944, 2021. https://doi.org/10.1109/JSTARS.2021.3059991
  7. Q. He, et. al., "SAR Target Recognition Based on Task-Driven Domain Adaptation Using Simulated Data," IEEE Geoscience and Remote Letters, Vol. 19, No. 4019205, 2022.
  8. A. Criminisi, et. al., "Region Filling and Object Removal by Exemplar-Based Image Inpainting," IEEE Trans. Image Processing, Vol. 13, No. 9, pp. 1200-1212, 2004. https://doi.org/10.1109/TIP.2004.833105
  9. A. Bochkovskiy, et. al., "YOLOv4 : Optimal Speed and Accuracy for Object Detection," 2020, arXiv: 2004:10934.
  10. D. Malmgren-Hansen, et. al., "Improving SAR Automatic Target Recognition Models with Transfer Learning from Simulated Data," IEEE Geoscience Remote Sensing Letters, Vol. 14, No. 9, pp. 1484-1488, 2017. https://doi.org/10.1109/LGRS.2017.2717486