라즈베리파이와 Yolo를 이용한 운전 도우미 시스템 설계 = Designing a Driver Assistant System Using Raspberry Pi and Yolo

다국어 초록 (Multilingual Abstract)

We propose a driver-assistance application designed to enhance road safety through real-time situational awareness.
The system provides contextual notifications such as the operating hours of bus-only lanes based on the vehicle’s
current geographic position, and issues warnings when jaywalking pedestrians or pedestrians appearing during
right-turn maneuvers are detected. The application employs a Raspberry Pi 4 equipped with a Pi Camera2 to
capture and stream forward-facing video from the vehicle. Object detection is performed using a YOLO-based
model, and the resulting detection metadata is transmitted to a central server. The server processes these data to
evaluate predefined traffic and safety conditions. When a potentially hazardous situation is identified, the server
generates an appropriate text-to-speech (TTS) message and transmits it back to the client application for immediate
auditory feedback to the driver.