• Annual Conference on Human and Language Technology
• /
• 1992.10a
• /
• pp.225-235
• /
• 1992

본 논문은 오프라인 필기체 한글 인식에 관한 연구로서, 입력 문자 영상에 대한 위치 정규화, 외곽선 추적 및 세선화의 전처리 과정을 거쳐 외곽선의 방향 성분 분포, 세선화한 결과의 방향 성분 분포, 구조적 특징점 분포 등의 특징을 추출한 다음, 획을 추출하여 획의 방향과 길이에 대한중점 분포 특징으로 정합하는 필기 한글의 인식을 위한 획 정합 방법을 제안하였다. 인식 시간의 단축을 위해 먼저 외곽선의 방향성분분포를 이용하여 대분류하였으며, 한글 사용 빈도수 상위 520자로 구성되는 필기 데이타에 대한 실험 결과, 평균 91%의 인식률과 평균 0.46초의 문자당 인식 시간을 보임으로써 제안된 획 정합 방법이 입력 문자의 잡영이나 획의 기울기에 대한 변형을 효과적으로 흡수할 수 있음을 알 수 있었다.

• Proceedings of the KIPE Conference
• /
• 2016.11a
• /
• pp.65-66
• /
• 2016

이 논문은 자성체 주변에 자기장과 자기장 기울기를 형성하여 자성체의 위치와 각도를 제어하는 자기장 제어 시스템을 위한 제어법을 제안한다. 자기장 제어 시스템을 구성하는 각 코일은 전류 제한 아래 동작하게 되는데 만약 각 코일이 독립적으로 제한되는 경우 합성된 자기장이 주어진 지령과 달라져 자성체의 정렬 및 병진운동 방향이 잘못 제어될 수 있다. 이 논문은 이를 방지하여 방향 우선 제어를 수행하기 위한 비례 전류 지령 보정법을 제안한다. 이와 함께 전압 제한으로 인해 발생할 수 있는 자성체의 과도기 방향 제어 오류를 방지하기 위한 PI 전류제어기 이득 설정 방법을 제안한다. 제안된 전류 지령 보정법의 유효성은 실험을 통해 확인되었다.

• Journal of the Korean Vacuum Society
• /
• v.15 no.4
• /
• pp.410-420
• /
• 2006

The dynamic effects, such as the steering and the screening effects during deposition on an epitaxial growth is studied by kinetic Monte Carlo simulation. In the simulation, we incorporates molecular dynamic simulation to rigorously take the interaction of the deposited atom with the substrate atoms into account, We find three characteristic features of the surface morphology developed by grazing angle deposition: (1) enhanced surface roughness, (2) asymmetric mound, and (3) asymmetric slopes of mound sides, Regarding their dependence on both deposition angle and substrate temperature, a reasonable agreement of the simulated results with the previous experimental ones is found. The characteristic growth features by grazing angle deposition are mainly caused by the inhomogeneous deposition flux due to the steering and screening effects, where the steering effects play the major role rather than the screening effects. Newly observed in the present simulation is that the side of mound in each direction is composed of various facets instead of all being in one selected mound angle even if the slope selection is attained, and that the slope selection does not necessarily mean the facet selection.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.8
• /
• pp.2017-2024
• /
• 1997

The main focus of this paper is the recognition of printed Hangul documents in terms of typeface, character size and character slope for IICS(Intelligent Image Communication System). The fixed-size blocks extracted from documents are analyzed in frequency domain for the typeface classification. The vertical pixel counts and projection profile of bounding box are used for the character size classification and the character slope classification, respectively. The MLP with variable hidden nodes and error back-propagation algorithm is used as typeface classifier, and Mahalanobis distance is used to classify the character size and slope. The experimental results demonstrated the usefulness of proposed system with the mean rate of 95.19% in typeface classification. 97.34% in character size classification, and 89.09% in character slope classification.

• Journal of Korea Water Resources Association
• /
• v.31 no.4
• /
• pp.363-373
• /
• 1998

In this study, random flow field in a nonstationary porous formation is characterized through cross covariances of the velocity with the log conductivity and the head. The hydraulic head and the velocity in saturated aquifers are found through stochastic analysis of a steady, two-dimensional flow field without recharge. Expression for these cross covariances are obtained in quasi-analytic forms all in terms of the parameters which characterize the nonstationary conductivity field and the average head gradient. The cross covariances with a Gaussian correlation function for the log conductivity are presented for two particular cases where the trend is either parallel or perpendicular to the mean head gradient and for separation distances along and across the mean flow direction. The results may be of particular importance in transport predictions and conditioning on field measurements when the log conductivity field is suspected to be nonstationary and also serve as a benchmark for testing nonstationary numerical codes. Keywords : cross covariance, nonstationary conductivity field, saturated aquifer, stochastic analysis.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2017.05a
• /
• pp.825-831
• /
• 2017

In this study, we perform the design of dome geometry for the composite pressure vessel with applying the equation of Fulton and Vasiliev considering external load(thrusts). Variables of the dome geometry are opening radius ratio(${\rho}_0$) from 0.1 to 0.5 and thrust level from 40kN to 200kN. We conduct Finite Element Analysis(FEA) by using ABAQUS. As a result, the strain of the composite pressure vessel has shown strain gradient from inner to outer of dome surface. And the strain gradient may cause crack of resin inside the composite laminate. Strain gradient of Fulton dome is monotonously decreased as the ${\rho}_0$ increases, but the strain gradient of Vasiliev dome bas shown some different trend. when ${\rho}_0{\leq}0.1$, strain gradient of Fulton's is higher than Vasiliev's. But when 0.1<${\rho}_0$<0.35, strain gradient of Vasiliev's becomes higher than Fulton's. And in the case of $0.35{\leq}{\rho}_0$, strain gradient of Vasiliev's is higher than Fulton's. So the Vasiliev dome is more effective in ${\rho}_0{\leq}0.1$ condition and Fulton dome is more effective in $0.35{\leq}{\rho}_0$ condition. So, it's important for dome design to consider the crack of resin cause of the strain gradient.

• Journal of Science Education
• /
• v.33 no.2
• /
• pp.193-206
• /
• 2009

The purpose of this study was to examine the effect of the consistent presentation of illustrations about the aligning direction of the Axis on the middle school students' acquisition and retention of astronomical concepts. This study was taken using the nonequivalent control-group pretest-posttest design on 116 7th middle school subjects. The same teaching and learning activities were given to both the experimental (n=59) and control groups (n=57) through three lessons. The experimental group was given a consistent presentation of the illustrations about the aligning direction of the Axis, while the control group was given an inconsistent presentation of the same illustrations about aligning direction. Two days after the three lessons, the 1st posttest was administered to compare the statistical difference of mean of both groups, using ANCOVA test. The result of ANCOVA test implicated that the consistent presentation of the illustrations about the aligning direction of the Axis had a positive influence on the experimental group's acquisition of the concepts. The 2nd posttest result for retention effect was given two month later by one-paired t-Test in each group and showed that the method had a positive effect on the experimental group, compared with control group. The results of this study implicated that paying careful attention to using the consistent illustration is highly beneficial for students' meaningful learning on astronomical concepts.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10b
• /
• pp.374-376
• /
• 1999

본 논문은 연속적인 영상에서 움직이는 물체의 광류를 예측하는데 있어서 웨이브릿과 에피폴라 평면(Epipolar-Plane Images, EPI)을 사용하여 물체의 움직임을 분석하는 기법을 제안한다. 하나의 영상에서 물체의 특징을 찾는데 2차 웨이브릿 변환이 사용되고 있으나 연적인 영상에 대한 분석에서 시간에 대한 또 하나의 변환을 해 줌으로서 움직이는 물체의 정보를 얻어낼 수 있다. 3차 웨이브릿 변환에서 유도된 데이터를 가지고 만들어지는 에피폴라 평면은 움직이는 물체의 광류를 예측하는데 있어서 기존의 방법보다 시간적으로 절약을 할 수 있다. 특히 서로 다른 방향에 대한 민감성을 보여 주는 웨이브릿 계수들은 움직이는 물체의 광류 예측에 많은 도움을 주고 있다. EPI 분석에서는 물체가 깊이 방향으로 움직이는 경우에도 물체의 각 에지들(edges)의 기울기를 분석함으로서 깊이 방향의 광류를 측정할 수 있다. 본 논문에서 제시한 3차 웨이브릿 변환과 EPI 기법의 조합으로 분석된 실험 결과와 그 전 연구들과의 비교가 마지막 부분에 서술되었다.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.6
• /
• pp.37-48
• /
• 2012

Image registration is a process to establish the spatial correspondence between the images of same scene, which are acquired at different view points, at different times, or by different sensors. In this paper, we propose an effective registration method for images acquired by multi-sensors, such as EO (electro-optic) and IR (infrared) sensors. Image registration is achieved by extracting features and finding the correspondence between features in each input images. In the recent research, the multi-sensor image registration method that finds corresponding features by exploiting NMI (Normalized Mutual Information) was proposed. Conventional NMI-based image registration methods assume that the statistical correlation between two images should be global, however images from EO and IR sensors often cannot satisfy this assumption. Therefore the registration performance of conventional method may not be sufficient for some practical applications because of the low accuracy of corresponding feature points. The proposed method improves the accuracy of corresponding feature points by combining the gradient orientation as spatial information along with NMI attributes and provides more accurate and robust registration performance. Representative experimental results prove the effectiveness of the proposed method.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2017.10a
• /
• pp.243-244
• /
• 2017

This paper introduces about the difference between the tunneling currents in the 1D and 2D directions for the calculation of the band-to-band tunneling currents of the tunneling field effect transistors. In the two-dimensional tunneling, diagonal tunneling is not calculated in the one-dimensional tunneling so that more accurate tunneling current can be calculated. Simulation results show that the tunneling in the two - dimensional direction has no effect on the voltage above the threshold voltage, but it affects the subthreshold swing below the threshold voltage.