• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.05b
• /
• pp.579-582
• /
• 2003

본 논문은 열화된 영상에서 문자 패턴의 자동 분할을 위해 농담정규화상관(NGC)법과 다중 템플릿을 이용하는 알고리즘을 제안한다. 기존의 NGC를 사용하는 검사 알고리즘은 환경조건의 영향으로 검사 영상의 획득이 불완전하다면 정합의 부독율(rejection rate)이 높아진다. 다중 템플릿의 상관관계를 이용하는 제안된 방법은 가시화가 졸지 않은 경우에도 문자부와 배경부를 효과적으로 분할하며, 이러한 방법을 실제 자동화 공정에서 획득된 영상을 이용하여 제안된 알고리즘을 적용하는 것을 목표로 한다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2008.05a
• /
• pp.169-173
• /
• 2008

본 연구는 저수량 지역 빈도분석(regional low flow frequency analysis)을 수행하기 위하여 일반최소자승법(ordinary least squares method)을 이용한 Bayesian 다중회귀분석을 적용하였으며, 불확실성측면에서의 효과를 탐색하기 위하여 Bayesian 다중회귀분석에 의한 추정치와 t 분포를 이용하여 산정한 일반 다중회귀분석의 추정치의 신뢰구간을 비교분석하였다. 각 재현기간별 비교결과를 보면 t 분포를 이용하여 산정된 평균 추정치와 Bayesian 다중회귀분석에 의한 평균 추정치는 크게 다르지 않았다. 그러나 불확실성 측면에서 평가해볼 때 신뢰구간의 상한추정치와 하한추정치의 차이는 Bayesian 다중회귀분석을 사용한 경우가 기존 방법을 사용한 경우보다 훨씬 작은 것으로 나타났으며, 이로부터 저수량(low flow) 지역 빈도분석을 수행하는 경우 Bayesian 다중회귀분석이 일반 회귀분석보다 불확실성을 표현하는데 있어서 우수하다는 결과를 얻을 수 있었다. 또한 낙동강 유역에 2개의 미계측 유역을 선정하고 구축된 Bayesian 다중회귀모형을 적용하여 불확실성을 포함한 미계측 유역에서의 저수량(low flow)을 추정하였으며 이와 같은 방법이 미계측 유역에서의 저수(low flow) 특성을 나타내는 데 있어서 효과적일 수 있음을 입증하였다.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2014.02a
• /
• pp.350.1-350.1
• /
• 2014

분포 브래그 반사기(distributed Bragg reflector; DBR)는 광센서, 도파로, 태양전지, 반도체 레이저 다이오드, 광검출기와 같은 고성능 광 및 광전소자 응용분야에 널리 사용되고 있다. 일반적으로, DBR은 박막의 두께를 4분의 1 파장(${\lambda}/4$)으로 가지는 서로 다른 저굴절율 물질과 고굴절율 물질을 교대로 적층 (pair)한 다중 pair로 제작되어지며, DBR의 반사 특성과 반사대역폭은 두 물질의 굴절율 차이와 pair의 수에 영향을 받는다. 그러나, 서로 다른 굴절율을 갖는 두 물질을 이용하는 DBR의 경우, 두 물질간 열팽창계수의 불일치, 접착력 문제, 높은 굴절율 차이를 갖는 물질 선택의 어려움 등 많은 문제점을 지니고 있다. 최근, 경사입사각증착법을 이용한 동일 재료(예, 인듐 주석 산화물, 게르마늄, 실리콘)기반의 DBR 제작 및 특성에 대한 연구가 보고되고 있다. 높은 입사각을 갖고 박막이 증착될 경우, 저율을 갖는 다공박막 제작이 가능하여 경사입사각증착법으로 homogeneous 물질 기반의 고반사 특성을 갖는 다중 pair의 DBR을 제작할 수 있다. 본 실험은, 갈륨비소 기판 위에 경사입사각증착법 및 전자빔증착법을 이용하여 중심파장 960 nm가 되는 이산화 티타늄 기반의 DBR을 제작하였고, 제작된 샘플의 증착된 박막의 표면 및 단면의 프로파일은 주사전자현미경을 사용하여 관찰하였으며, UV-Vis-NIR 스펙트로미터를 이용하여 반사율 특성을 조사하였다.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.9 no.5
• /
• pp.640-647
• /
• 1998

Microstrip patch antennas are analyzed by a multiresolution wavelet method. The spectral Green's dyad of the structure is obtained and its joint spatial-spectral domain representations are presented. Based on the joint spatial-spectral domain representation, we show that the spectral-domain wavelets are useful in the analysis of this problem. We obtain the matrix equations of the integral equations of this Green's dyad by using the method of moment(MoM), and efficiently solve the problem using the spectral domain wavelet transform concepts in conjuction with the conjugate gradient method. The results for a single-layered square patch are compared with those of conventional MoM and CG-FFT.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.2
• /
• pp.277-287
• /
• 2013

It is important to detect the gene-gene interaction in GWAS (genome-wide association study). There have been many studies on detecting gene-gene interaction. The one is D-MDR (dummy multifoactor dimensionality reduction) method. The goal of this study is to evaluate the power of D-MDR for identifying gene-gene interaction by simulation. Also we applied the method on the identify interaction effects of single nucleotide polymorphisms (SNPs) responsible for economic traits in a Korean cattle population (real data).

• The Korean Journal of Applied Statistics
• /
• v.26 no.3
• /
• pp.549-568
• /
• 2013

This paper discusses conversion methods between multiple decrement models and associated single decrement models. One of most popular assumptions on fractional age is UDD(uniform distribution of decrement) or constant force of mortality in actuarial practice. Instead of these assumptions, this paper suggests cubic spline interpolation to approximate the distribution of fractional age with the continuous force of decrements. Conversion formulas are derived. The comparisons of these two methods based on the numerical data show that the cubic spline interpolation approach is more accurate.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.16 no.2
• /
• pp.235-242
• /
• 1998

This study aims to detect and to reduce the multipath errors which are main errors source in high precious surveying such as GPS-aided aerial triangulation and Car Navigation. which reference receivers being fixed, when kinematic receivers move continuously, multipath is performed using smoothed code measurement and pure code measurement in the network. Through this methods, 3D RMS errors are reduced into about 30% in the single differential code solution to the kinematic receiver. This is based on the fact that the network adjustment are performed at multiple reference receivers, but positioning is carried out by the single differential methods between a reference receiver and a kinematic receiver. So it was supposed that this methods reduced the correlation errors including the atmospheric errors using the nearest receivers and can be mixed with another methods.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 1999.11a
• /
• pp.204-209
• /
• 1999

In general, neural networks are widely used for the category classification. But when low contrast images, such as multispectral images, are used as the input of neural networks, neural networks converge very slowly and are of bad performance. To overcome this problem, we propose a new image enhancement method which consists of smoothing process, finding the main valley and enhancement process. And the enhanced images by the proposed method are used as the input of neural networks for the category classification. When the new category classification method is applied to multispectral LANDSAT TM images, we verified that neural networks converge very fast and overall category classification performance is improved.

• Journal of the Microelectronics and Packaging Society
• /
• v.29 no.1
• /
• pp.67-70
• /
• 2022

In the present work, we have designed to optimize the optical filter structures of the 1-dimensional photonic quasicrystals (1D PQCs) characteristic for the COVID-19 UV sterilization. The simulator using MATLAB program and ourselves manufacturing calculation codes. After making the aperiodic (and complexed) multi-layer structure model, we establish the transfer matrix method (TMM) for model by the operator conversion. By the using the MATLAB, we derive a matrix for the designed complexed multi-layer structure by applying the equations to the model by obtaining the reflectance and transmittance from the matrix. We also prove the possibility of application in optical filter for UV sterilization.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.27 no.1
• /
• pp.135-140
• /
• 2003

The convergence of finite volume method (FVM) or discrete ordinate method (DOM) is known to degrade for optical thickness greater than unity and large scattering albedo. The present article presents a convergence acceleration procedure for the FVM based on a full approximation storage (FAS) multigrid method. Among a variety of multigrid cycles, the V-cycle is used and the full multigrid algorithm (FMG) is applied to an analysis of radiation in irregular two-dimensional geometry. Solution convergence is discussed for the several cases of various optical thickness and scattering albedo. At small scattering albedo and optical thickness, there is no advantage to using the multigrid method for calculation CPU time. For large scattering albedo greater than 0.5 and optical thickness greater than unity, however, the multigrid method improves the convergence and the solution is rapidly obtained.