• 한국컴퓨터정보학회논문지
• /
• 제14권11호
• /
• pp.73-79
• /
• 2009

본 논문에서는 Local Binary Pattern 히스토그램의 템플릿 매칭을 이용한 얼굴 표정 인식에서 인식률을 높이는 방법을 제안한다. 이 방법에서, 주어진 얼굴 영상은 작은 크기의 블록으로 분할되고 각 블록에서 구해진 LBP 히스토그램은 블록 특징으로 사용된다. 입력 영상에서의 블록 특징과 모델의 해당블록 특징 사이에서 블록 상이도가 계산된다. 주어진 영상과 모델 영상 사이의 영상 상이도는 블록 상이도의 가중 합으로 계산된다. 기존의 방법들에서는 직관에 따른 블록 가중치를 사용하는데 본 논문에서는 블록 가중치를 트레이닝 샘플들로부터 최적화를 통해서 구하는 방법을 제안하고 있다. 실험을 통해서 제안된 방법이 기존의 방법보다 우수함을 보인다.

• 정보처리학회논문지B
• /
• 제13B권7호
• /
• pp.643-652
• /
• 2006

의사결정 트리는 데이터 마이닝의 분류와 예측 작업에 주로 사용되는 기법 중의 하나이다. 실생활에서 공간의사결정을 위한 분류를 수행할 때에는 인접 데이터의 위치와 분산도를 고려하는 것이 매우 중요하다. 기존의 공간 의사결정 트리는 데이터의 공간적 특성을 표현하기 위해 각 객체간의 유클리디안 거리비율을 엔트로피로 반영하여 트리 구축 시 이용하였다. 그러나 이것은 공간 객체간의 거리 비율만을 설명할 뿐 공간 차원에서의 데이터 분산 정도와 각 분류된 클래스간의 연관관계 등은 파악할 수 없다는 한계점이 있었다 본 논문에서는 분산도와 차별도 기반의 공간 엔트로피를 이용하여 공간 데이터의 분포도를 반영하는 공간 의사결정 트리를 제안한다 분산도는 분류된 클래스 내의 공간 객체 분포도를 나타내고 차별도는 다른 클래스 내 공간 객체와의 분포도 및 관계성을 나타낸다. 이러한 분산도와 차별도의 비율을 엔트로피 계산 시 이용함으로써 비공간적 속성으로 분류된 각 클래스가 공간적으로는 얼마나 뚜렷하게 분류되는지 알 수 있게 한다. 제안 기법은 정확성과 계산 비용에 있어서 기존 기법보다 각각 약 18%, 11%의 성능 향상을 보였다.

• 한국콘크리트학회:학술대회논문집
• /
• 한국콘크리트학회 1998년도 가을 학술발표논문집(II)
• /
• pp.597-602
• /
• 1998

The composite systems, consisting of R/C Columns-Steel Beams, are reasonable structures because of their constructional and economical advantages, workability and so on. But, it is difficult to apply the composite systems to actual design due to material dissimilarity and complicate stress flow in the connection. This study aims to propose the hybrid beam-column connection system with structural tee and through experimental research make clear the shear and moment resistance capacity and stress transfer mechanism.

• 한국정보디스플레이학회:학술대회논문집
• /
• 한국정보디스플레이학회 2008년도 International Meeting on Information Display
• /
• pp.718-721
• /
• 2008

Novel computational integral imaging technique with enhanced depth sensitivity is proposed. For each lateral position at a given depth plane, the dissimilarity between corresponding pixels of the elemental images is measured and used as a suppressing factor for that position. Experimental and simulation results show that reconstructed depth image on the incorrect depth plane is effectively suppressed.

• Journal of Information Display
• /
• 제11권2호
• /
• pp.68-75
• /
• 2010

The relation between light sources and screen cells is considered part of the theoretical model of an autostereoscopic 3D display. The geometry of the image and observer regions is presented, including the cases of single and multiple regions. The characteristic function is introduced. Formulas for the geometric parameters are obtained, including areas and angles. Special attention is drawn to the screen location. The method of transforming the formulas between regions is stated. For multiple regions, geometric dissimilarity was found. This allows the model to be applied in finding the geometric characteristics of multiview and integral-imaging 3D displays.

• Journal of the Korean Data and Information Science Society
• /
• 제21권6호
• /
• pp.1353-1359
• /
• 2010

Basic objective in cluster analysis is to discover natural groupings of items. In general, clustering is conducted based on some similarity (or dissimilarity) matrix or the original input data. Various measures of similarities between objects are developed. In this paper, we consider a clustering of huge categorical real data set which shows the aspects of time-location-activity of Korean people. Some useful similarity measure for the data set, are developed and adopted for the categorical variables. Hierarchical and nonhierarchical clustering method are applied for the considered data set which is huge and consists of many categorical variables.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 하계학술대회 논문집 G
• /
• pp.2243-2245
• /
• 1998

In this paper, we treat image segmentation as a graph partitioning problem. and use the normalized cut for segmenting the graph. The normalized cut criterion measures both the total dissimilarity between the different graphs as well as the total similarity within the groups. The minimization of this criterion can formulated as a generalized eigenvalues problem. We have applied this approach to segment static image. This criterion can be shown to be computed efficiently by a generalized eigenvalues problem

• Genomics & Informatics
• /
• 제14권4호
• /
• pp.181-186
• /
• 2016

Glucose tolerance tests have been devised to determine the speed of blood glucose clearance. Diabetes is often tested with the standard oral glucose tolerance test (OGTT), along with fasting glucose level. However, no single test may be sufficient for the diagnosis, and the World Health Organization (WHO)/International Diabetes Federation (IDF) has suggested composite criteria. Accordingly, a single multi-class trait was constructed with three of the fasting phenotypes and 1- and 2-hour OGTT phenotypes from the Korean Association Resource (KARE) project, and the genetic association was investigated. All of the 18 possible combinations made out of the 3 sets of classification for the individual phenotypes were taken into our analysis. These were possible due to a method that was recently developed by us for estimating genomic associations using a generalized index of dissimilarity. Eight single-nucleotide polymorphisms (SNPs) that were found to have the strongest main effect are reported with the corresponding genes. Four of them conform to previous reports, located in the CDKAL1 gene, while the other 4 SNPs are new findings. Two-order interacting SNP pairs of are also presented. One pair (rs2328549 and rs6486740) has a prominent association, where the two single-nucleotide polymorphism locations are CDKAL1 and GLT1D1. The latter has not been found to have a strong main effect. New findings may result from the proper construction and analysis of a composite trait.

• Journal of Plant Biotechnology
• /
• 제31권1호
• /
• pp.19-23
• /
• 2004

본 연구는 붓꽃속 식물의 형태적 특성 조사와 RAPD분석을 통해 유전적 다양성과 근연성을 밝히고 유전자 도입을 위한 기초적인 자료를 얻기 위하여 실시하였다. RAPD를 이용하여 국내에서 수집한 자생붓꽃종의 DNA다형성을 조사한 결과 40개의 primer중 8개의 유용한primer가 선발되었고 이들 primer는 108개의 밴드를 나타내었다. 이들 중 다형성 밴드는 107 개였다. Jaccard (1-S$_{J}$ )법에 의한 비유사도 계수는 0.095∼0.699로 나타났고, dendrogram에서 나타난 유전적 근연성은 크게 세 개의 군으로 분류되었다. I군은 연미붓꽃, II군은 타래붓꽃, 제비붓꽃, 노랑꽃창포, 노랑무늬붓꽃이 속해 있었으며, III군은 금붓꽃, 각시붓꽃, 붓꽃, 꽃창포가 속해 있었다.

• 대한산업공학회지
• /
• 제41권1호
• /
• pp.43-49
• /
• 2015

The aim of this study was to model unusual association with declarative knowledge by positive affect using ACT-R cognitive architecture. Existing research related with cognitive modeling tends to pay a lot of attention to strong and negative cognitive moderator. Mild positive affect, however, has far-reaching effects on problem solving and decision making. Typically, subjects with positive affect were more likely to respond to unusual associates in a word association task than subjects with neutral affect. In this study, a cognitive model using ACT-R cognitive architecture was developed to show the effect of positive affect on the cognitive organization related with memory. First, we organized the memory structure of stimulus word 'palm' based on published results in a word association task. Then, we decreased an ACT-R parameter that reflects the amount of weighting given to the dissimilarity between the stimulus word and the associate word to represent reorganized memory structure of the model by positive affect. As a result, no significant associate probability difference between model prediction and existing empirical data was found. The ACT-R cognitive architecture could be used to model the effect of positive affect on the unusual association by decreasing (manipulating) the weight of the dissimilarity. This study is useful in conducting model-based evaluation of the effects of positive affect in complex tasks involving memory, such as creative problem solving.