국내기사
휴대용 FMCW 레이더의 호흡 신호를 활용한 AI 기반 인원 탐지 성능 향상 연구 = AI-based enhancement of personnel detection performance using respiration signals from a portable FMCW radar
This study evaluates the personnel detection performance of a 24 GHz FMCW radar (HLK-LD2450) and aims to enhance the detection of concealed or motionless individuals by incorporating respiration-based features. Unlike previous studies focused on medical environments, this work constructed and utilized Human and Non-human datasets collected under conditions simulating real battlefield disturbances such as shielding, foliage, wind, and rain. Statistical and respiration-related features extracted from time-series radar data were applied to a CNN model. The results show improved detection accuracy and stability even in complex environments, demonstrating that respiration signals play a key role in enhancing the reliability of radar-based personnel detection.