메뉴 건너뛰기
소속 기관 / 학교 인증
인증하면 논문, 학술자료 등을  무료로 열람할 수 있어요.
한국대학교, 누리자동차, 시립도서관 등 나의 기관을 확인해보세요
(국내 대학 90% 이상 구독 중)
고객센터 ENG
주제분류

논문 기본 정보

저자정보
(국민대학교) (국민대학교)
저널정보
한국자동차공학회 한국자동차공학회논문집 한국자동차공학회논문집 제31권 제8호
오류 신고하기

피인용 2

검색

    초록·키워드

    In this paper, an improved Mono-SLAM(Simultaneous Localization and Mapping) algorithm is proposed to address the depth estimation of a monocular camera without using additional sensors. The proposed SLAM uses a depth estimation based on a deep-learning approach, and generates a refined depth map through semantic segmentation. The proposed algorithm is validated through the RGB-D mode of ORB SLAM2, and it demonstrated that localization accuracy can be improved, compared to using the Mono version of ORB SLAM2. Moreover, it can utilize RGB-D cameras in outdoor environments, though it may lead to performance degradation. To solve performance degradation in the depth estimation of the depth map at the boundary between distant objects due to insufficient ground information, a refined depth map that uses semantic segmentation is adopted, thus improving localization accuracy.

    최근 본 자료 전체보기

      UCI(KEPA) : I410-ECN-0102-2023-556-001945467