http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
박규도(Gyu-Do Park),강수혁(Soo-Hyeok Kang) 대한전자공학회 2023 대한전자공학회 학술대회 Vol.2023 No.6
In recent years, the manufacturing and logistics industries have been utilizing artificial intelligence technology to improve productivity. Among them, predicting potential risks and detecting them in advance to prevent collisions in product shipping areas is crucial. In this paper, we developed a detection system based on a deep learning model to prevent forklift safety accidents. To this end, driving videos of forklifts were collected at the product shipping site and a model was built that recognizes forklifts, trucks, and people by learning them. Using this model, a system was implemented that sounds an alarm to the driver and pedestrian when there is a possibility of collision between the forklift or truck and the person. The detection system can be monitored in real-time with Mean AP 55.3%, 13.2 FPS with VGG16-SSD model and Mean AP 45.9%, 17.9 FPS with MobileNetV2-SSD model. This system will contribute to creating a safe working environment.
MobileNet을 이용한 마스크 착용 유무 판별 및 체온 측정 시스템 개발
박규도(Gyu-Do Park ),김근수(Geun-Su Kim),강수혁(Soo-Hyeok Kang) 대한전자공학회 2020 대한전자공학회 학술대회 Vol.2020 No.8
Coronavirus Disease 2019 (COVID-19) has widely spread all over the world since the beginning of 2020. It is a highly contagious disease that can lead to acute respiratory difficulties or multiple organ failure in severe cases. In an effort to prevent community infection, most large facilities have managers to monitor thermal imaging and recommend wearing face masks on entry and exit. This paper introduces a system that determines whether a face mask is worn or not with deep learning machine vision and obtains thermal information and notifies the user simultaneously.
김근수(Geun-Su Kim),박규도(Gyu-Do Park),강수혁(Soo-Hyeok Kang) 대한전자공학회 2020 대한전자공학회 학술대회 Vol.2020 No.8
In this study, a fire detection system using images in an embedded environment was implemented using the YOLO algorithm. In a single controller, the neural network computation speed is very slow, so a neural network computation accelerator must be used to improve the computation speed. In this paper, Intel"s NCS1 was used as an accelerator, and it deals with how to implement an embedded system.
차량 배기가스 인증모드 무인주행을 위한 차속 추종 자동화 시스템 개발
강수혁(Soo-Hyeok Kang),김대우(Dae-Woo Kim),성봉진(Bong-Jin Seoug),김영학(Yeong-Hak Kim),박규도(Gyu-Do Park),김근수(Geun-Su Kim) 대한전자공학회 2020 대한전자공학회 학술대회 Vol.2020 No.8
In this paper, we propose a vehicle speed tracking automation system for autonomous driving in the vehicle emission certification mode. We controlled the acceleration by creating an acceleration voltage, and the deceleration control used a brake robot. Using this system, the exhaust gas certification mode test can be automated. The proposed system was tested through actual vehicle driving.