• Aerospace Engineering and Technology
• /
• v.9 no.2
• /
• pp.36-43
• /
• 2010

A GEO satellite launched by Arian 5 ECA launcher will be located in a transfer orbit where it requires several Apogee burn maneuvers to reach the target orbit. To obtain the required performance of Apogee burn maneuvers, a calibration of gyro drift error needs to be performed before each maneuver. In this paper, a unique gyro calibration scheme which is applied to COMS is described and the calibration error budget analysis is performed.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.06a
• /
• pp.293-295
• /
• 2013

최근 3D 비디오의 흥행과 더불어 사람들이 스테레오스코픽 비디오를 쉽게 접할 수 있게 되었다. 하지만 부적절한 스테레오 스코픽 비디오는 시청자로 하여금 어지럼증, 눈의 피로 등을 느끼게 할 수 있다. 따라서 비디오 컨텐츠를 분석하여 피로도를 예츨할 수 있는 지표의 개발이 필요하다. 스테레오스코픽 비디오에서 피로도를 유발시키는 원인은 여러 가지가 있으며, 본 논문에서는 비디오의 카메라 움직임 정보에 집중한다. 카메라의 흔들림은 심각한 멀미를 유발할 수 있다. 이는 스테레오스코픽 비디오에서 더욱 강렬하게 나타나게 된다. 우리는 스테레오스코픽 비디오의 흔들림을 측정하기 위해 3D 복원 기술을 사용한다. 특히, 좌안 비디오와 우안 비디오의 정보를 모두 활용하는 최적화된 알고리즘을 제안한다. 이를 통해 우리는 각 프레임의 카메라 위치와 방향 정보를 획득할 수 있다. 마지막으로 획득한 3차원 카메라 움직임 정보로부터 카메라의 흔들림을 잘 표현할 수 있는 점수를 계산한다.

• The Magazine of the IEIE
• /
• v.33 no.1 s.260
• /
• pp.35-45
• /
• 2006

최근 휴대폰에서 개인 정보 보안의 중요성이 대두되고 있으며, 이에 따라 생체인식 기능을 내장한 휴대폰에 관심이 집중되고 있으므로, 본 논문에서는 휴대용 기기에 홍채인식기술을 적용하기 위한 방법을 제안하고 하다. 기존의 홍채인식 알고리즘은 인식을 위해 확대된 홍채영상을 사용하여 처리하였고, 이러한 홍채영상을 획득하기 위해서 고 배율의 줌 렌즈(zoom lens)와 초점 렌즈(focus lens)를 사용하였다. 그런데, 휴대폰에서의 홍채인식을 시도하기 위해 줌렌즈와 초점렌즈를 장착하게 되면 가격이 상승하고 부피가 증가되어, 작고 휴대하기 편리한 휴대폰의 특징에 맞지 않아 사용하기 어려운 문제가 있다. 그러나 최근 휴대폰의 멀티미디어 기기 융복합 추세로 인해 휴대폰 내에 장착된 메가 픽셀 카메라(Mega-pixel Camera)의 성능이 급속히 발전함에 따라, 고 배율의 줌, 초점 렌즈 없이도 확대된 홍채영상의 획득이 가능하게 되었다. 즉, 메가 픽셀 카메라 폰을 사용하여 사용자로부터 원거리에서 취득한 얼굴영상에서 홍채 영역이 홍채인식을 위해 충분한 픽셀정보를 가지게 된다. 그러나 이러한 경우 입력 얼굴영상에서 눈 영역을 먼저 찾는 과정이 요구된다. 본 논문에서는 이러한 얼굴영상에서 각막에 의해 반사되는 조명 반사광을 기반으로 휴대폰에서의 홍채인식을 위한 고속 동공검출 방법을 제안한다. 또한 입력 영상 신호를 디지털 신호로 변환하는 과정에서 밝기와 대조 값을 조정하여 동공의 검은 영역과 조명 밝은 반사 위치를 추출함으로써, 정확한 홍채 영역을 보다 빠르고 쉽게 추출할 수 있는 방법 역시 제안한다. 그리고 일반적으로 휴대폰에서 홍채 인식의 경우 손으로 들고 사용하므로, 손 흔들림에 의한 영상 흐림 현상이 빈번하게 발생하게 된다. 이러한 문제를 해결하기 위하여 본 연구에서는 영상 복원 기법을 적용하여 흐려진 홍채 영상을 복원하는 기법을 제안한다. 마지막으로, 휴대폰의 경우 실외에서 사용이 빈번함으로, 입력 홍채 영상에서 태양광에 의한 영향을 많이 받게 된다. 이러한 문제를 해결하여 홍채 인식 성능을 개선할 수 있는 방법 역시 소개하고자 한다.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.50 no.9
• /
• pp.617-625
• /
• 2022

This study investigates the use of movable calibration target for gyroscopic and accelerometer bias compensation of inertial measurement units for firing lane alignment. Calibration source is detected with the help of vision sensor and its information in fused with other sensors on launcher for error correction. An algorithm is proposed and tested in simulation. It has been shown that it is possible to compensate sensor biases in firing launcher in few seconds by accurately estimating the location of calibration target in inertial frame of reference.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.50 no.9
• /
• pp.639-646
• /
• 2022

Landing site searching algorithms are developed for a quadrotor using a depth map in unknown environment. Guidance and control system of Unmanned Aerial Vehicle (UAV) consists of a trajectory planner, a position and an attitude controller. Landing site is selected based on the information of the depth map which is acquired by a stereo vision sensor attached on the gimbal system pointing downwards. Flatness information is obtained by the maximum depth difference of a predefined depth map region, and the distance from the UAV is also considered. This study proposes three landing methods and compares their performance using various indices such as UAV travel distance, map accuracy, obstacle response time etc.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1302-1308
• /
• 2019

This paper presents an approach for classifying image made by acquired position data from a 2D Lidar sensor with a convolutional neural network (CNN). Lidar sensor has been widely used for unmanned devices owing to advantages in term of data accuracy, robustness against geometry distortion and light variations. A CNN algorithm consists of one or more convolutional and pooling layers and has shown a satisfactory performance for image classification. In this paper, different types of CNN architectures based on training methods, Gradient Descent(GD) and Levenberg-arquardt(LM), are implemented. The LM method has two types based on the frequency of approximating Hessian matrix, one of the factors to update training parameters. Simulation results of the LM algorithms show better classification performance of the image data than that of the GD algorithm. In addition, the LM algorithm with more frequent Hessian matrix approximation shows a smaller error than the other type of LM algorithm.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.9
• /
• pp.1786-1792
• /
• 2007

In this paper, we proposed a recognition system for resident registration cards using ART-1 and PCA algorithm. To extract registration numbers and issue date, Sobel mask and median filter are applied first and noise removal follows. From the noise-removed image, horizontal smearing is used to extract the regions, which are binarized with recursive binarization algorithm. After that vortical smearing is applied to restore corrupted lesions, which are mainly due to the horizontal smearing. from the restored image, areas of individual codes are extracted using 4-directional edge following algorithm and face area is extracted by the morphologic characteristics of a registration card. Extracted codes are recognized using ART-1 algorithm and PCA algorithm is used to verify the face. When the proposed method was applied to 25 real registration card images, 323 characters from 325 registration numbers and 166 characters from 167 issue date numbers, were correctly recognized. The verification test with 25 forged images showed that the proposed verification algorithm is robust to detect forgery.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.4
• /
• pp.752-758
• /
• 2012

SNS(Social Networking Services) is oneline service that enable users to construct human network through their relation on web, such as following relation, friend relation, and etc. Recently, owing to the advent of digital devices (smart phone, tablet PC) which embedded GPS some applications which provide services with spatial relevance and social relevance have been released. Such an online service is called LBSNS. It is required to use spatial filtering so as to build the LBSNS system that enable users to subscribe information of interesting area. For spatial filtering, user and tweet attaches location information which divide into static property presenting fixed area and dynamic property presenting user's area changed along the moving user. In the case of using a location information including dynamic property, Continuous query occurred from the moving user causes the problem in server. In this paper, we propose spatial filtering algorithm using Virtual Grid for reducing frequency of query, and conclude that frequency of query on using Virtual Grid is 93% decreased than frequency of query on not using Virtual Grid.

• Journal of the Korea Society of Computer and Information
• /
• v.19 no.5
• /
• pp.43-52
• /
• 2014

In this paper, we propose Gravity Removal and Vector Rotation algorithm for counting steps of wearable device, and we evaluated the proposed GRVR algorithm with Micro-Electro-Mechanical (MEMS) 3-axis accelerometer equipped in low-power wearable device while the device is mounted on various positions of a walking or running person. By applying low-pass filter, the gravity elements are canceled from acceleration on each axis of yaw, pitch and roll. In addition to DC-bias removal and the low-pass filtering, the proposed GRVR calculates acceleration only on the yaw-axis while a person is walking or running thus we count the step even if the wearable device's axis are rotated during walking or running. The experimental result shows 99.4% accuracies for the cases where the wearable device is mounted in the middle and on the right of the belt, and 91.1% accuracy which is more accurate than 83% of commercial 3-axis pedometer when worn on wrist for the case of axis-rotation.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.5
• /
• pp.88-96
• /
• 2014

In modern electronic warfare systems, the necessity of a more accurate estimation method based on non-AOA (arrival of angle) measurement, such as TDOA and FDOA, have been increased. The previous researches using single TDOA have been carried out in terms of not only the development of emitter location algorithms but also the enhancement of measurement accuracy. Recently, however, the combined TDOA/FDOA method is of considerable interest because it is able to estimate the velocity vector of a moving emitter and acquire a pair of TDOA and FDOA measurements from a single sensor pair. In this circumstance, it is needed to derive the required FDOA measurement accuracy in order that the TDOA/FDOA combined localization system outperforms the previous single TDOA localization systems. Therefore, we analyze the contribution of FDOA measurement accuracy to emitter location, then propose the criterion based on CRLB (Cramer-Rao lower bound). Simulations are included to examine the validity of the proposed criterion by using the Gauss-Newton algorithm.