A Photonic Radar Aided DCNN-Based Classification of LSS Targets Using their ISAR-Images: DIAT-ISAR-sATIs

International Journal of Electronics and Communication Engineering

© 2025 by SSRG - IJECE Journal

Volume 12 Issue 11

Year of Publication : 2025

Authors : Nargis Akhter, Ajay Waghumbare, A. Arockia Bazil Raj

10.14445/23488549/IJECE-V12I11P108

How to Cite?

Nargis Akhter, Ajay Waghumbare, A. Arockia Bazil Raj, "A Photonic Radar Aided DCNN-Based Classification of LSS Targets Using their ISAR-Images: DIAT-ISAR-sATIs," SSRG International Journal of Electronics and Communication Engineering, vol. 12,  no. 11, pp. 103-114, 2025. Crossref, https://doi.org/10.14445/23488549/IJECE-V12I11P108

Abstract:

Protective measures for national, military, or civil surveillance systems are critical for the automatic detection and identification of various aerial Low-altitude Slow-speed Smaller size (LSS) targets accurately. In this study, a diversified “Defense Institute of Advanced Technology-Inverse Synthetic Aperture Radar-small Aerial Target Images (DIAT-ISAR-sATIs)” dataset consisting of 4320 Inverse Synthetic Aperture Radar (ISAR) images of five distinct aerial LSS targets is constructed using an ultra-broadband (6-12 GHz Instantaneous Bandwidth (IBW)) Stepped Frequency Modulated Continuous Wave (SFMCW) photonic radar. After data augmentation, the dataset was increased to 6000 samples. Additionally, a Deep Convolutional Neural Network (DCNN) approach based on transfer learning was introduced for the classification of aerial LSS targets. The method achieved a classification accuracy of 98.67% on validation samples and 95.56% on open-field samples by utilizing a pretrained MobileNetV2 as the feature extractor, with minimal occurrences of false negatives and false positives. The experimental classes investigated in this study are 1) an RC plane; 2) a mini-helicopter; 3) a DJIF 450; 4) a Scythe 4s racer; 5) a bionic-bird; and 6) a combination of a mini-helicopter and a bionic-bird. Experimental results demonstrate that the proposed pre-trained DCNN model outperforms existing pre-trained models on the DIAT-ISAR-sATIs dataset.

Keywords:

Photonic radar, Inverse synthetic aperture radar imaging, Low-altitude slow-speed smaller-size targets, Deep convolutional neural network.

References:

[1] Samedh Sachin Kari, A. Arockia Bazil Raj, and K. Balasubramanian, “Evolutionary Developments of Today’s Remote Sensing Radar Technology-Right From the Telemobiloscope: A Review,” IEEE Geoscience and Remote Sensing Magazine, vol. 12, no. 1, pp. 67-107, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[2] Hossein Emami et al., “Simultaneous Echo Power and Doppler Frequency Measurement System Based on Microwave Photonics Technology,” IEEE Transactions on Instrumentation and Measurement, vol. 66, no. 3, pp. 508-513, 2017.
[CrossRef] [Google Scholar] [Publisher Link]
[3] Bir Bhanu, “Automatic Target Recognition: State of the Art Survey,” IEEE Transactions on Aerospace and Electronic Systems, vol. AES-22, no. 4, pp. 364-379, 1986.
[CrossRef] [Google Scholar] [Publisher Link]
[4] Massimiliano Pieraccini, Lapo Miccinesi, and Neda Rojhani, “RCS Measurements and ISAR Images of Small UAVs,” IEEE Aerospace and Electronic Systems Magazine, vol. 32, no. 9, pp. 28-32, 2017.
[CrossRef] [Google Scholar] [Publisher Link]
[5] Sampurna De, and A.A. Bazil Raj, “A Survey on Photonics Technologies for Radar Applications,” Journal of Optics, vol. 52, pp. 90-119, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[6] Andrea Manno-Kovacs et al., “Image Based Robust Target Classification for Passive ISAR,” IEEE Sensors Journal, vol. 19, no. 1, pp. 268-276, 2018.
[CrossRef] [Google Scholar] [Publisher Link]
[7] Hari Kishan Kondaveeti, and Valli Kumari Vatsavayi, “Abridged Shape Matrix Representation for the Recognition of Aircraft Targets from 2D ISAR Imagery,” Advances in Computational Sciences and Technology, vol. 10, no. 5, pp. 1103-1122, 2017.
[Google Scholar] [Publisher Link]
[8] Yann LeCun, Yoshua Bengio, and Geoffrey Hinton, “Deep Learning,” Nature, vol. 521, pp. 436-444, 2015.
[CrossRef] [Google Scholar] [Publisher Link]
[9] Xueru Bai et al., “REMI: Few-Shot ISAR Target Classification via Robust Embedding and Manifold Inference,” IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 4, pp. 6000- 6013, 2025.
[CrossRef] [Google Scholar] [Publisher Link]
[10] Chow Yii Pui et al., “Target Classification for 3D-ISAR Using CNNs,” IEEE Transactions on Aerospace and Electronic Systems, vol. 60, pp. 94-105, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[11] A. Lazarov, and C. Minchev, “ISAR Image Recognition Algorithm and Neural Network Implementation,” Cybernetics and Information Technologies, vol. 17, no. 4, pp. 183-199, 2017.
[Google Scholar] [Publisher Link]
[12] Harish C. Kumawat et al., “DIAT-μSAT: Small Aerial Targets’ Micro-Doppler Signatures and Their Classification Using CNN,” IEEE Geoscience and Remote Sensing Letters, vol. 19, pp. 1-5, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[13] Dongsuk Park et al., “Radar-Spectrogram-Based UAV Classification Using Convolutional Neural Networks,” Sensors, vol. 21, no. 1, pp. 1-18, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[14] Neeraj Pandey, and Shobha Sundar Ram, “Classification of Automotive Targets Using Inverse Synthetic Aperture Radar Images,” IEEE Transactions on Intelligent Vehicles, vol. 7, no. 3, pp. 675-689, 2022.
[CrossRef] [Google Scholar] [Publisher Link]
[15] Arthur Holland Michel, “Counter-Drone Systems,” Center for the Study of the Drone at Bard College, Report, pp. 1-23, 2018.
[Google Scholar] [Publisher Link]
[16] Ian Goodfellow, Yoshua Bengio, and Aaron Courville, Deep Learning, MIT Press, pp. 1-775, 2016.
[Google Scholar] [Publisher Link]
[17] M.N. Saidi et al., “Aircraft Target Recognition: A Novel Approach for Features Extraction from ISAR Images,” 2009 International Radar Conference "Surveillance for a Safer World" (RADAR 2009), Bordeaux, France, pp. 1-5, 2009.
[Google Scholar] [Publisher Link]
[18] José Capmanyet al., “Microwave Photonics: Current Challenges towards Widespread Application,” Optics Express, vol. 21, no. 9, pp. 22862-22867, 2018.
[CrossRef] [Google Scholar] [Publisher Link]
[19] Caner Ozdemir, Inverse Synthetic Aperture Radar Imaging With MATLAB Algorithms, 2nd ed., General & Introductory Electrical & Electronics Engineering, John Wiley & Sons, pp. 1-672, 2021.
[Google Scholar] [Publisher Link]
[20] Donald R. Wehner, High-Resolution Radar, 2nd ed., Artech House, Boston, pp. 1-593, 1994.
[Google Scholar] [Publisher Link]
[21] Chun-Ting Lin et al., “Optical Millimeter-Wave Signal Generation Using Frequency Quadrupling Technique and No Optical Filtering,” IEEE Photonics Technology Letters, vol. 20, no. 12, pp. 1027-1029, 2008.
[CrossRef] [Google Scholar] [Publisher Link]
[22] Valli Kumari Vatsavayi, and Hari Kishan Kondaveeti, “Efficient ISAR Image Classification Using MECSM Representation,” Journal of King Saud University - Computer and Information Sciences, vol. 30, no. 3, pp. 356-372, 2018.
[CrossRef] [Google Scholar] [Publisher Link]
[23] P. Anantha Prabha, G. Suchitra, and R. Saravanan, “Cephalopods Classification Using Fine Tuned Lightweight Transfer Learning Models,” Intelligent Automation & Soft Computing, vol. 35, no. 3, pp. 3065-3079, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[24] Ruihang Xue et al., “Sequential ISAR Target Classification Based on Hybrid Transformer,” IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-11, 2022.
[CrossRef] [Google Scholar] [Publisher Link]
[25] Soyoon Park et al., “Classification of ISAR Images Generated from Model-based Approaches Using Deep Learning,” 2025 International Conference on Information Networking (ICOIN), Chiang Mai, Thailand, pp. 312-316, 2025.
[CrossRef] [Google Scholar] [Publisher Link]
[26] Xingyu He, Ningning Tong, and Xiaowei Hu, “Automatic Recognition of ISAR Images Based on Deep Learning,” 2016 CIE International Conference on Radar (RADAR), Guangzhou, China, pp. 1-4, 2016.
[CrossRef] [Google Scholar] [Publisher Link]