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
(충북대학교) (충북대학교)
저널정보
제어로봇시스템학회 제어로봇시스템학회 논문지 제어로봇시스템학회 논문지 제27권 제8호
오류 신고하기

피인용 0

검색

    초록·키워드

    Metric segmentation based defect detection method is a method of detecting defects using metric learning. In the supervised learning-based method, overfitting of the convolutional neural network occurs in the environment of a small amount of defect data set, which makes it difficult to detect defects. To solve this, we propose a defect detection system based on metric learning. The metric vector of the input image is extracted through the metric embedding model. The center of the normal metric vector is obtained through l2 norm of the extracted metric vector. The defect is detected based on the center of the obtained normal metric vector. Experimental results prove that the proposed method has higher detection accuracy than the existing method in an environment with few training data.

    최근 본 자료 전체보기

      UCI(KEPA) : I410-ECN-0101-2021-003-001917543