ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper The Design of CNN Hardware IP to Classify a Bird and Drones in Radar System
Cited - time in scopus Share share facebook twitter linkedin kakaostory
Authors
YoungSeok Baek, BonTae Koo
Issue Date
2023-02
Citation
International Conference on Electronics, Information and Communication (ICEIC) 2023, pp.426-428
Language
English
Type
Conference Paper
Abstract
In this paper, we presented a CNN structure and RTL IP implementation that detects drones and birds, which are disadvantages of existing drone detection radar systems, using micro-Doppler-based micro-motion. CNN architecture was designed with RTL IP. The RTL IP was verified with 500 dwell test scenario data, and the Matlab reference model and RTL results all matched except for 4 data, achieving 99% performance. And, it shows the FPGA implementation HW results of Drone Detection Radar System and CNN hardware IP.