http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
한철수(Han Cheol-su),박병욱(Park Byoung-uk),서재갑(Suh Jae-gap),김학윤(Kim Hack-yoon) 한국정보기술학회 2007 Proceedings of KIIT Conference Vol.2007 No.-
인간은 좌우 두 귀 청취로, 한정적이지만 음원의 방위각뿐만 아니라 앙각에 대해서도 그 정보를 추출 가능하다. 즉, 인간의 청각 기능은 비교적 높은 레벨의 잡음 환경하에 있어서도 회화를 유지하기 위한 중요한 요소이다. 따라서 인간은 두 귀 청취에 의한 음원 방향 추정과 그에 따른 음원 분리가 행해지고 있다. 본 논문에서는 주파수 영역 두 귀 청취 모델을 이용한 음원의 도래 방향 추정 방법에 관하여 논한다. 이 방법은 두 귀간 위상차와 두 귀간 레벨차를 이용하여 주파수 대역에서 음원의 방향을 추정하는 방법이다. 또한, 이 방법을 입증하기 위하여 일정 위치에 음원을 생성 후, 컴퓨터 시뮬레이션을 수행한 결과, 양호한 방향 추정 결과를 얻을 수 있었다.
수정된 MobileNet을 이용한 과일의 결점 분류 알고리즘
소민혁(Min-Hyeok So),한철수(Cheol-Su Han),김학윤(Hack-Yoon Kim) 한국정보기술학회 2020 한국정보기술학회논문지 Vol.18 No.7
The presence of one or a few defective fruits in a fruit package reduces the commercial value of the fruit. Currently, defected fruits are being sorted by naked eyes. However, the lack of objectivity by the conventional sort method and decrease in farming population are becoming problems. Various image processing methods have been introduced to overcome these problems. The convolutional neural network (CNN), an enhanced image processing method, provides an efficient solution for sorting the defective fruits. Recently, MobileNet has been proposed to incorporate CNN in a device with low arithmetic performance. Nevertheless, when it is used to sort defective fruits, there is a possibility of identifying good fruits as defected ones. Therefore, the thesis proposes a revised MobileNet algorithm that can be applied to a high-speed fruit sort machine that is being used in the industry as of now. Moreover, it is to prove the usefulness of the proposed algorithm through a performance evaluation.
근적외선을 이용한 실시간 과일의 당도 측정 시스템 구현
소민혁(Min-Hyeok So),한철수(Cheol-Su Han),김학윤(Hack-Yoon Kim) 한국정보기술학회 2020 한국정보기술학회논문지 Vol.18 No.3
With the increase in income, the consumer‘s standard on fruit selection is changing from quantity to quality. Among them, the soluble solids content(SSC), which represents the ratio of sugars contained in fruits, greatly affects the taste, making it the most important factor for consumer’s fruit preferences. A representative procedure of accurate SSC measurement uses the refracting saccharometer. However, this method holds drawbacks to the research procedure for it causes damage to fruit in the process of measurement. To resolve this problem, a non-destructive measuring method has been proposed. Various methods that have been proposed mainly focuses on the algorithm which measures the SSC of fruits. Therefore, the researchers will apply the proposed system that can sort fruits under certain SSC standards onto the currently available electronic fruit sorting machines and provide a GUI (Graphic User Interface) control system. The performance evaluation of the proposed SSC measurement system verifies its accuracy and reproducibility. Moreover, the utility of the proposed system is proved by statistical methods.