• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2020년도 전력전자학술대회
• /
• pp.425-426
• /
• 2020

본 논문에서는 영구자석형 동기전동기의 벡터제어를 수행하는데 필요한 회전자의 위치, 특히 기동시의 초기 회전자 위치를 정확하게 검출하기 위하여 자기식 위치 센서를 사용하는 방법을 제안하고, 기존의 광학식 인크리멘탈 로터리 엔코더를 사용하는 경우에 비하여 검출 성능이 우수함을 확인하였다.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제31권4호
• /
• pp.403-410
• /
• 2004

바이모달 음성인식기는 잡음 환경하 음성인식 성능을 향상하기 위해 고안되었다. 바이모달 음 성인식기에 있어 영상을 통한 시각 특징 추출은 매우 중요한 역할을 하며 이를 위한 입술 위치 검출은 시각 특징 추출을 위한 중요한 선결 과제이다 본 논문은 색상분포와 SVM을 이용하여 시각 특징 추출을 위한 입술 위치 검출 방법을 제안하였다. 제안된 방법은 얼굴색/입술 색상 분포를 학습하여 이로부터 입술의 초기 위치를 빠르게 찾아내고 SVM을 이용하여 입술의 정확한 위치를 찾음으로써 정확하고 빠르게 입술의 위치를 찾도록 하였으며 실험을 통해 바이모달 인식기에 적용하기에 적합함을 알 수 있었다.

• 대한전자공학회논문지SP
• /
• 제39권3호
• /
• pp.23-35
• /
• 2002

시선 위치 추적이란 모니터상에 사용자가 쳐다보고 있는 지점을 파악해 내는 기술이다 이 논문에서는 컴퓨터 비젼 방법을 이용하여 사용자가 모니터 상에 어느 지점을 쳐다보고 있는지를 파악(시선 위치 추적)하는 새로운 방법을 제안한다. 시선위치를 파악하기 위해 본 논문에서는 얼굴 영역 및 얼굴 특징점(양 눈, 양 콧구멍, 입술 끝점 등)을 2차원 카메라 영상으로부터 추출하였으며, 이들의 움직임으로부터 카메라 보정 및 매개변수 추정 방법등을 이용하여 초기 3차원 위치를 추정해 내었다. 이후 모니터 상의 한 지점을 쳐다보기 위해 사용자가 얼굴을 움직이는 경우 이러한 얼굴의 3차원 움직임 량 역시 자동으로 추정하였다. 이로부터 변화된 얼굴 특징점의 3차원 위치를 계산해 낼 수 있었으며, 이를 바탕으로 모니터 상의 시선 위치를 구하였다. 실험 결과, 19인치 모니터상의 임의의 지점을 사용자가 쳐다보았을 때, 약 2.01인치의 시선 위치에러 성능을 얻었다.

• 한국정보통신학회논문지
• /
• 제13권7호
• /
• pp.1429-1436
• /
• 2009

본 논문에서는 다양한 조명하에서의 단일 얼굴 영상에 대해 유전자 알고리즘과 형판 정합을 이용하여 빠르게 눈동자를 검출하는 방법을 제안한다. 유전 알고리즘을 이용한 기존의 눈동자 검출 방법은 초기 개체군의 위치에 민감하여 낮은 눈 검출율을 보이며, 도한 그 결과가 일관적이지 않은 문제점을 갖는다. 이와 같은 문제점을 해결화기 위해 얼굴영상에서 지역적 최소치를 추출하고 형판과 가장 높은 적합도를 가지는 개체들로 초기 개체군을 생성 하였다. 각각의 개체는 형판의 기하학적 변환 정보로 구성되며, 형판 정합에 의해 눈동자가 검출된다. 실험을 통하여 본 논문에서 제안한 눈 후보 검출을 통하여 단일 영상에서도 눈 검출의 정확도와 높은 검출률을 확인하였다.

• Journal of Electrical Engineering and Technology
• /
• 제13권4호
• /
• pp.1515-1527
• /
• 2018

This paper presents a sensorless speed control of IPMSM (Interior Permanent Magnet Synchronous Motor) using the high-frequency (HF) square wave injection method. In the proposed HF pulsating square wave injection method, injection voltage is applied into the estimated d-axis of rotor and high-frequency induced q-axis current is considered to estimate the rotor position. Conventional square wave injection methods may need complex demodulation process to find rotor position, while in the proposed method, an easy demodulation process based on the rising-falling edge of the injected voltage and carrier induced q-axis current is implemented, which needs less processing time and improves control bandwidth. Unlike some saliency-based sensorless methods, the proposed method uses maximum torque per ampere (MTPA) strategy, instead of zero d-axis command current strategy, to improve control performance. Furthermore, this paper directly uses resultant d-axis current to detect the magnet polarity and eliminates the need to add an extra pulse injection for magnet polarity detection. As experimental results show, the proposed method can quickly find initial rotor position and MTPA strategy helps to improve the control performance. The effectiveness of the proposed method and all theoretical concepts are verified by mathematical equations, simulation, and experimental tests.

• 반도체디스플레이기술학회지
• /
• 제19권2호
• /
• pp.11-15
• /
• 2020

In a typical initial setup of a PCB component inspection system, operators should manually input various information such as category, position, and inspection area for each component to be inspected, thus causing much inconvenience and longer setup time. Although there are many deep learning based object detectors, RetinaNet is regarded as one of best object detectors currently available. In this paper, a method using an extended RetinaNet is proposed that automatically detects its component category and position for each component mounted on PCBs from a high-resolution color input image. We extended the basic RetinaNet feature pyramid network by adding a feature pyramid layer having higher spatial resolution to the basic feature pyramid. It was demonstrated by experiments that the extended RetinaNet can detect successfully very small components that could be missed by the basic RetinaNet. Using the proposed method could enable automatic generation of inspection areas, thus considerably reducing the setup time of PCB component inspection systems.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2015년도 추계학술발표대회
• /
• pp.1073-1075
• /
• 2015

In this paper vision based system for detection of golf ball position and spin velocity in screen golf game is introduced. Firstly, an algorithm is presented for finding ball position and its initial linear velocity on the golf court, and secondly, golf ball spin is calculated. These parameters are vital for simulator in order to simulate a real golf environment. Finally, these algorithm have been implemented with acceptable result.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2003년도 추계학술대회 논문집
• /
• pp.119-122
• /
• 2003

This paper describes the sensorless technique for the surface-mounted permanent-magent synchronous motor(SPMSM or PMSM) drive based on magnetic saliency. The control technique is a sensorless control algorithm that injects the high frequency voltage to the stator terminal in order to estimate the rotor position and speed. The rotor position and speed for sensorless vector control is achieved by appropriate signal processing to extract the position information from the stator current in the low speed range including zero speed. Proposed sensorless algorithm using the double-band hysteresis controller and initial rotor position detection exhibits excellent reference tracking and increased robustness. Experimental results are presented to verify the feasibility of the proposed control schemes.

• International Journal of Contents
• /
• 제9권3호
• /
• pp.9-14
• /
• 2013

In this paper, we will examine single face tracking algorithms with scaling function in a mobile device. Face detection and tracking either in PC or mobile device with scaling function is an unsolved problem. Standard single face tracking method with particle filter has a problem in tracking the objects where the object can move closer or farther from the camera. Therefore, we create an algorithm which can work in a mobile device and perform a scaling function. The key idea of our proposed method is to extract the average of skin color in face detection, then we compare the skin color distribution between the detected face and the tracking face. This method works well if the face position is located in front of the camera. However, this method will not work if the camera moves closer from the initial point of detection. Apart from our weakness of algorithm, we can improve the accuracy of tracking.

• 한국수정란이식학회지
• /
• 제30권4호
• /
• pp.335-340
• /
• 2015

The objective of this study was to evaluate the initial detection time and development of the fetal genital structures using ultrasound in twelve pregnant small bitches. The initial detection time of the fetal genital structures was as follows: genital tubercle at days 32.6; os penis at days 45.2; labia at days 45.7; scrotum at days 47.5. Ultrasonograms of fetal genital structure according to gestational stage were as follows: Undifferentiated stage (before day 35), the genital tubercle was observed to have a small elevation and just a hyper-echogenic structure in the midline between the umbilical cord and the tail in male and female fetus. Migration stage (between day 35~45), the genital tubercle was observed as a hyper-echogenic, bilobular, oval shaped and the genital tubercle began to migrate from the initial position toward the umbilical cord in males, and toward the tail in females. Differentiated stage (after day 46), the penis and os penis were observed to stand out in the abdominal wall and the scrotum was observed toward the perineal region in male fetuses. The labia was detected at the base of the tail in female fetuses. These results indicate that ultrasound of fetal genital structures could be useful for fetal gender determination and a completely prepartum evaluation of the canine fetus.