Recent advancements in automotive technology along with sustained investments in road infrastructure and associated transportation facilities have significantly reduced the frequency of road traffic accidents. Nevertheless, human-related factors, particularly driver negligence, continue to account for most traffic incidents. To address accidents caused by driver inattention and drowsiness, this paper proposes a digitalcode- based traffic safety sign system capable of delivering direct and real-time information to drivers through onboard vision-recognition devices, such as vehicle dashcams and black-box systems, which are now widely deployed in modern vehicles. Several candidate digital codes applicable to traffic safety signs were reviewed to develop the proposed system. Among them, the ArUco3 marker was selected owing to its simple geometric structure and superior recognition performance in camera-based detection environments. To experimentally validate the applicability of the proposed system, a prototype mechanism was implemented, in which a vehicle-mounted vision-recognition device detected traffic safety signs embedded with ArUco3 markers and subsequently provided preconfigured warning and guidance information to the driver in real time through audiovisual outputs. Performance evaluation experiments focusing on recognition rate, which is considered the most critical factor governing the practical applicability of the proposed system, were conducted on a real-world test track under actual driving conditions. To ensure a rigorous evaluation, test specimens were fabricated with dimensions equivalent to those of conventional traffic safety signs while varying the fundamental cell size of the embedded ArUco3 markers. Additionally, repeated driving experiments were performed by incrementally increasing the vehicle speed from 30 to 150 km/h. The experimental results demonstrated that the vehiclemounted vision-recognition system successfully detected and recognized the proposed digital traffic safety signs, even under high-speed driving conditions. These findings verify the effectiveness and practical feasibility of the proposed ArUco3-marker-based digital traffic safety sign system. Furthermore, in future heterogeneous traffic environments where autonomous and human-driven vehicles will coexist, the proposed system is expected to provide an additional channel for acquiring road information, thereby enabling more effective responses to unexpected incidents and emergencies.

ABSTRACT
1. 서론
    1.1. 연구의 배경 및 목적
    1.2. 연구 내용 및 수행 절차
2. 도입 필요성 및 기존사례 고찰
    2.1. 도입 필요성
    2.2. 사례 조사 및 착안사안
3. 디지털 코드형 교통안전표지 제안
    3.1. ArUco3 마커의 활용
    3.2. ArUco3 마커 안전표지 설계
    3.3. ArUco3 마커 안전표지 제작과 활용
4. ArUco3 마커 안전표지 인식률 성능 평가
    4.1. 실험용 안전표지 제작
    4.2. 실험용 영상인식 장치
    4.3. 실험 절차
    4.4. 실험 결과
5. 결론 및 향후과제
REFERENCES

• 강성인(한국도로공사 도로교통연구원 수석연구원) | Kang Sungin
• 이기영(한국도로공사 도로교통연구원 연구위원) | Lee Kiyoung
• 강성환((주)다인랩 대표이사) | Kang Sunghwan
• 이성관 (국립군산대학교 조교수) | Lee Seongkwan Mark Corresponding author