A5033538854- Advanced SAR Imaging Techniques
- Radar Systems and Signal Processing
- Target Tracking and Data Fusion in Sensor Networks
- Hand Gesture Recognition Systems
- Inertial Sensor and Navigation
- Machine Fault Diagnosis Techniques
- Optical Wireless Communication Technologies
- Full-Duplex Wireless Communications
- Bladed Disk Vibration Dynamics
- CRISPR and Genetic Engineering
- Viral Infectious Diseases and Gene Expression in Insects
- Impact of Light on Environment and Health
- Microwave Imaging and Scattering Analysis
- COVID-19 diagnosis using AI
- Animal Genetics and Reproduction
- Radio Wave Propagation Studies
- Gait Recognition and Analysis
- Non-Invasive Vital Sign Monitoring
- Machine Learning in Healthcare
- Terahertz technology and applications
- Spectroscopy Techniques in Biomedical and Chemical Research
- Video Surveillance and Tracking Methods
- Advanced Chemical Sensor Technologies
- Diverse Topics in Contemporary Research
- Identification and Quantification in Food
Daegu Gyeongbuk Institute of Science and Technology
2019-2024
This paper presents development of 24GHz millimeter wave smart radar for intelligent street lighting system. The developed operates at a frequency with 200MHz bandwidth and CW (continuous wave) mode. two radars are used to cover the in both directions detect obstacles moving speed more than 1 km/h including pedestrians. detection controls lighting. Therefore, energy-saving performance has been improved because proposed system works only when around streetlight exist.
Recently, Doppler radar-based foot gesture recognition has attracted attention as a hands-free tool. for various gestures is still very challenging. So far, no studies have yet dealt deeply with of based on radar and deep learning model. In this paper, we propose method using new high-compression signature image learning. By means AlexNet model, created by extracting dominant features via Singular Value Decomposition (SVD) processing; four different including kicking, swinging, sliding,...
This study investigates the feasibility of an automatic detection system for foreign objects in food using terahertz radar and deep learning techniques. The experimental setup comprises a source, beam splitters collimators, receiver with 32 x 32, 1024-pixel area scanner. Received images representing signal strength transmitted through noodles serve as input model after preprocessing to eliminate noise. A binary decision is then made on whether contains or not.
The signal transmitted from radar is not reflected a single point when the by complex target. Resultantly, amplitude and phase of received can be changed because target has lots scatterers. changes mean Glint RCS, respectively. Although RCS that caused same scatters are uncorrelated, however, they independent completely. Therefore, this paper proposes method for generating using random number generator. And time correlations respectively implemented in frequency domain each power spectral...
추적 레이더에서 각 오류는 정확한 위치 추적을 방해하는 요소이다. 오류를 발생시키는 원인으로는 열잡음, 서보 잡음, 글린트가 있다. 사용되는 칼만 필터는 대체로 잡음환경을 부가 백색 가우시안 잡음인 열잡음으로 가정하여 동작한다. 그러나 본 논문에서 고려하고자하는 글린트는 열잡음과 다른 특성을 갖기 때문에 필터로 충분히 제거할 수 있는지 확인할 필요가 논문에서는 글린트의 전력밀도함수를 활용하여 시간영역에서의 상관특성을 결정하고, 통계적 특성은 분포와 라플라시안 분포가 혼합된 형태로 모델링하였다. 대표적인 추적레이더인 모노펄스 영향을 확인한 후, 한국형 위성발사체인 KSLV-Ⅰ 경로에 적용시켰을 때 글린트를 있는지를 확인하였다. 결과적으로, 확장 칼만필터는 저주파 성분을 추정하여 RMS값으로 약 20%정도 감소시킨다는 것을 있었다.
In this paper, an integrated mid-range automotive radar system with both Blind Spot Detection (BSD) and Lane Change Assistance (LCA) functions was developed. The system's maximum detection range extended to 80m update time of 50ms or less. To assess the performance in real-world driving conditions, complex dynamic scenarios, including lane changing following, straight curved road environments were tested at proving ground. developed Blind-Spot evaluated determine its accuracy detecting...
This paper proposes a structured pruning-network architecture search (NAS) algorithm for lightweight deep-learning radar foot gesture recognition in conventional models to quantitatively evaluate its performance. Our goal is recognize gestures using CW radar, generate their STFT unique signatures, and build system that could be implemented on an edge device. The proposed scheme shows model size FLOPs were reduced, sub-optimal device based MobileNet was obtained with slight decrease accuracy.
Purpose: With the increasing national interest in health, number of health examination centers is growing rapidly, and it as independent medical institutes separated from hospitals. functions size institutes, considerations for testees, who are most important users centers, have taken back seat. In particular, programs that take on a sequential traffic line, to be aware space each room, but lack scientific evaluation method this has resulted great discomforts testees using center. Method:...
Pneumonia is a common disease with frequent hospitalization and bacteremic pneumonia associated high mortality rate. And comorbidity one of the most important determinants mortality. Our study evaluated prognostic significance Charlson's weighted index comorbidities (WIC) in patients pneumonia. A retrospective based on an analysis medical records 767 who had positive blood culture admission university-affiliated tertiary care hospital between 2010 - 2017. Their mean WIC was 1.7±2.8, It...
본 논문에서는 CW(Continuous Wave) 레이더 시스템에서 캡스트럼을 적용하여 드론의 회전하는 블레이드를 검출하는 방법을 제안한다. 드론은 기존 항공기에 비해 크기가 작아 수신 신호의 전력을 측정하는 방법으로는 검출하기 어려우므로, 블레이드 회전에 의해 발생한 마이크로 도플러 특성을 활용하는 방법이 사용된다. 대표적인 검출 STFT(Short Time Fourier Transform) 방식이 있다. STFT 방식은 반복적으로 신호를 누적하는 과정이 필요하므로 긴 연산 시간이 요구되는 단점이 STFT의 단점을 극복하기 위해 캡스트럼 탐지한다. 우선, 블레이드의 단면적을 모델링하고, 이를 통해 포함하는 신호에 적용하였다. 최종적으로, 성능 평가를 수행된 시뮬레이션 결과로부터, 존재 여부를 임계치를 쉽게 알 수 있으며, 회전 속도에 대한 정보를 얻을