Journal
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
Volume 18, Issue 11, Pages 1946-1950Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/LGRS.2020.3011114
Keywords
Radar imaging; Drones; Radar polarimetry; Airborne radar; Radar cross-sections; Image color analysis; Convolutional neural network (CNN); micro-Doppler signature (MDS); radar signal analysis; radar signal processing
Categories
Funding
- Chungnam National University
Ask authors/readers for more resources
The study proposed a drone classification method for polarimetric radar based on CNN and image processing methods, which improved accuracy by enhancing weak micro-Doppler features. A novel image structure for three-channel image classification CNN was introduced to utilize received polarimetric signal, along with an image processing method to reduce data size while maintaining high classification accuracy.
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.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available