닫기
Loading..

Please wait....

학술대회 프로시딩

홈 홈 > 연구문헌 > 학술대회 프로시딩 > 대한전자공학회 학술대회 > 2018년 대한전자공학회 추계학술대회

2018년 대한전자공학회 추계학술대회

Current Result Document : 269 / 334 이전건 이전건   다음건 다음건

한글제목(Korean Title) YOLO를 이용한 국내 교통 표지판 검출 및 인식
영문제목(English Title) Korean Traffic-Sign Detection and Recognition using YOLO
저자(Author) 공성익   이상설   장성준   최병호   Sungick Kong   Sang-Seol Lee   Sung-Joon Jang   Byeongho Choi  
원문수록처(Citation) VOL 41 NO. 02 PP. 0832 ~ 0834 (2018. 11)
한글내용
(Korean Abstract)
영문내용
(English Abstract)
Autonomous vehicle technology is evolving with deep learning. In particular, the traffic sign recognition technology is a technique that notifies the driver of the signboard information of the current driving to the driver when the driver can not recognize the signboard while driving or the signboard information is missing from the GPS database. In this paper, we collected data directly on the road to secure the road situation and traffic sign data in Korea. Using the collected data, we conducted an experiment to transmit traffic sign information to the driver through the real - time operating system using the YOLO network structure. Through the experimental results, we can find the necessary measures to apply to actual roads and identify the parts to improve the performance according to detection rate and operation speed.
키워드(Keyword)   
파일첨부 PDF 다운로드