http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
유선진,이상윤 한국통신학회 2010 韓國通信學會論文誌 Vol.35 No.12
Pose variation is a critical problem in face recognition. Three-dimensional(3D) face recognition techniques have been proposed, as 3D data contains depth information that may allow problems of pose variation to be handled more effectively than with 2D face recognition methods. This paper proposes a pose-normalized 3D face modeling method that translates and rotates any pose angle to a frontal pose using a plane fitting method by Singular Value Decomposition(SVD). First, we reconstruct 3D face data with stereo vision method. Second,nose peak point is estimated by depth information and then the angle of pose is estimated by a facial plane fitting algorithm using four facial features. Next, using the estimated pose angle, the 3D face is translated and rotated to a frontal pose. To demonstrate the effectiveness of the proposed method, we designed 2D and 3D face recognition experiments. The experimental results show that the performance of the normalized 3D face recognition method is superior to that of an un-normalized 3D face recognition method for overcoming the problems of pose variation.
유선진 ( Sun Jin You ),박준성 ( Joon Sung Park ),김지현 ( Jee Hyun Kim ),박수경 ( Su Kyoung Park ),배상철 ( Sang Cheol Bae ),김근호 ( Gheun Ho Kim ),강종명 ( Chong Myung Kang ),박문향 ( Moon Hyang Park ),이창화 ( Chang Hwa Lee 대한신장학회 2009 Kidney Research and Clinical Practice Vol.28 No.5
Purpose: Clinical treatment for lupus nephritis largely depends upon histological renal biopsy classification. But it has been reported that serologic biochemical markers are not strongly associated with pathologic classification. The aim of this study is to see whether serologic markers could predict pathologic class of lupus nephritis for appropriate treatment. Methods: We investigated 67 patients, who underwent renal biopsy with lupus nephritis at Hanyang University Hospital between January, 2005 and August, 2007. Biological markers for this study are hematuria, proteinuria, serologic data of lupus activity and azotemia. They were retrospectively analyzed from patients grouped by ISN/RPS 2003 lupus nephritis classification. Results: Total 67 patients (men 5, women 62) were enrolled and the mean age of the patients was 30.6±9 years. The number of patient group by pathologic classification was 4 cases for class II, 15 cases for class III, 30 cases for class IV and 15 cases for class V. Spot urine protein to creatinine ratio more than 3 increased in class IV group statistically (p=.007). C3 level decreased more in class IV group than class III, V groups. Ten patients showed azotemia, and 9 of them were class IV group (p=.048). Conclusion: The patients with more increased proteinuria, decreased C3 level and azotemia showed more frequently in class IV group. Hence those three biological markers may be a clinical clue to pathologic diagnosis.
유선진 ( Sunjin Yu ),고완기 ( Koh Wan Ki ),김상훈 ( Sang Hoon Kim ) 한국정보처리학회 2013 한국정보처리학회 학술대회논문집 Vol.20 No.2
본 논문에서는 다중 손 끝점 검출을 위해 특징 추출 기법 및 이를 기반으로한 손 끝점 검출 알고리즘을 제안한다. 특징 추출을 위해 Local Binary Feature(LBP)을 사용하였고 특징의 차원을 축소하기 위해 Principal Component Analysis(PCA) 기법을 이용하였다. 손 끝점 판별을 위해 Reduced multivariate polynomial Model(RM) Classifier를 사용하여 실험 결과 제안된 손 끝점 검출 기법이 다양한 환경에서 동작 하는 것을 확인 하였다.
유선진(Sunjin Yu),윤창용(Changyong Yoon) 한국지능시스템학회 2017 한국지능시스템학회논문지 Vol.27 No.4
본 논문은 이동 로봇용 안드로이드 기반 시스템에 적용할 수 있는 물체 추적 알고리즘을 제안한다. 안드로이드 시스템은 저가이면서 휴대가 간편하고 범용적인 목적을 위해 사용될 수 있다는 장점이 있다. 이를 위해 스마트폰 카메라로부터 입력된 영상안의 객체를 실시간으로 추적하는 파티클 필터 기반 알고리즘을 안드로이드 시스템에 구현한다. 제안된 알고리즘은 움직임 신호를 블루투스 통신을 위한 패킷 형태로 이동 로봇에게 송신함으로써 이동로봇이 움직이는 객체를 추적할 수 있도록 한다. 또한, 객체 추적시 움직이는 장애물이 발견되고 객체를 완전히 가리는 경우에 초음파 센서를 함께 이용하여 로봇의 이동을 잠시 멈춤으로써 이동로봇의 추적 정확도가 감소되지 않도록 한다. 실험결과에서는 다른 추적 알고리즘과 성능을 비교함으로써 본 시스템에 적합한 최적의 알고리즘을 제안한다. This paper proposes an object tracking algorithm that can be applied to Android based systems for mobile robots. The Android system has the advantage that it can be used for low cost, portable and general purposes. To accomplish this, we implement a particle filter based algorithm to track an object in the input images from a smartphone camera in real time on the Android system. The proposed algorithm transmits the motion signal to the mobile robot in packet type for Bluetooth communication so that the mobile robot can track a moving object. In addition, when a moving obstacle is detected and completely occlude an object, ultrasonic sensors are used together to pause the movement of the robot so that the tracking accuracy of the robot is not reduced. In the experimental results, we propose an optimal algorithm forh its system by comparing performance with other tracking algorithms.
기술보고 : 스파크 트리거에 의한 비전기식 뇌관의 기폭 시스템
유선진 ( Seon Jin Yu ),강대진 ( Dae Jin Kang ),김남수 ( Nam Soo Kim ),장형두 ( Hyong Doo Jang ),양형식 ( Hyung Sik Yang ) 대한화약발파공학회 2011 화약발파 Vol.29 No.1
전기적 충격에 안전한 비전기뇌관은 지하 굴착에 널리 이용되고 있다. 그러나 국내 많은 현장에서 경제적인 이유로 전용 격발기 대신 전기뇌관으로 대신하고 있는 경우가 많다. 스파크 트리거는 이러한 비전기뇌관의 특성을 활용하지 못하는 기폭 시스템에 의한 발파사고를 막기 위해 개발 되었다. 이 시스템은 비싼 튜브가 필요없어서 경제적이며 기폭 후에 플라스틱 폐기물이 남지 않아 환경 친화적인 것으로 판단된다. Non-electric detonator has been used in underground excavations because of its strong resistance against electric impacts. However, electric detonator is often used to initiate the non-electric detonator instead of using an exclusive non-electric blasting machine due to economical reason. Spark Trigger is introduced as a solution of unexpected explosive hazard from using an electric detonator as an initiator of non-electric system. Since Spark Trigger System does not need expensive tube and no plastic waste is left, this system is proved to be more economical and eco-friendly initiate system than the standard non-electric initiating system.
유선진(Sunjin Yu),김중락(Joongrock Kim),이상윤(Sangyoun Lee) 대한전자공학회 2007 대한전자공학회 학술대회 Vol.2007 No.7
Pose-variation factors present a significant problem in 2D face recognition. To solve this problem, there are various approaches for a 3D face acquisition system which was able to generate multi-view images. However, this created another pose estimation problem in terms of normalizing the 3D face data. This paper presents a 3D head pose-normalization method using 2D and 3D interaction. The proposed method uses 2D information with the AAM(Active Appearance Model) and 3D information with a 3D normal vector. In order to verify the performance of the proposed method, we designed an experiment using 2.5D face recognition. Experimental results showed that the proposed method is robust against pose variation.