Improved Drone Classification Using Polarimetric Merged-Doppler Images

• Published in: IEEE geoscience and remote sensing letters Vol. 18; no. 11; pp. 1946 - 1950
• Main Authors: Kim, Byung Kwan, Kang, Hyun-Seong, Lee, Seongwook, Park, Seong-Ook
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.11.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accuracy; Airborne radar; Anechoic chambers; Artificial neural networks; Classification; Continuous radiation; Convolutional neural network (CNN); Doppler sonar; Drones; Evaluation; Image classification; Image color analysis; Image processing; micro-Doppler signature (MDS); Neural networks; Polarimetric radar; Polarimetry; Radar; Radar cross-sections; Radar equipment; Radar imaging; Radar polarimetry; radar signal analysis; radar signal processing
• ISSN: 1545-598X; 1558-0571
• DOI: 10.1109/LGRS.2020.3011114

We propose a drone classification method for polarimetric radar, based on convolutional neural network (CNN) and image processing methods. The proposed method improves drone classification accuracy when the micro-Doppler signature is very weak by the aspect angle. To utilize received polarimetric signal, we propose a novel image structure for three-channel image classification CNN. To reduce the size of data from four different polarization while securing high classification accuracy, an image processing method and structure are introduced. The data set is prepared for a three type of drone, with a polarimetric Ku-band frequency modulated continuous wave (FMCW) radar system. Proposed method is tested and verified in an anechoic chamber environment for fast evaluation. A famous CNN structure, GoogLeNet, is used to evaluate the effect of the proposed radar preprocessing. The result showed that the proposed method improved the accuracy from 89.9% to 99.8%, compared with single polarized micro-Doppler image. We compared the result from the proposed method with conventional polarimetric radar image structure and achieved similar accuracy while having half of full polarimetric data.