• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권5호
• /
• pp.1766-1784
• /
• 2014

Finger vein images contain rich oriented features. Local line binary pattern (LLBP) is a good oriented feature representation method extended from local binary pattern (LBP), but it is limited in that it can only extract horizontal and vertical line patterns, so effective information in an image may not be exploited and fully utilized. In this paper, an orientation-selectable LLBP method, called generalized local line binary pattern (GLLBP), is proposed for finger vein recognition. GLLBP extends LLBP for line pattern extraction into any orientation. To effectually improve the matching accuracy, the soft power metric is employed to calculate the matching score. Furthermore, to fully utilize the oriented features in an image, the matching scores from the line patterns with the best discriminative ability are fused using the Hamacher rule to achieve the final matching score for the last recognition. Experimental results on our database, MMCBNU_6000, show that the proposed method performs much better than state-of-the-art algorithms that use the oriented features and local features, such as LBP, LLBP, Gabor filter, steerable filter and local direction code (LDC).

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제16권2호
• /
• pp.104-110
• /
• 2016

Dense local image descriptors like SIFT are fruitful for capturing salient information about image, shown to be successful in various image-related tasks when formed in bag-of-words representation (i.e., histograms). In this paper we consider to utilize these dense local descriptors in the object tracking problem. A notable aspect of our tracker is that instead of adopting a point estimate for the target model, we account for uncertainty in data noise and model incompleteness by maintaining a distribution over plausible candidate models within the Bayesian framework. The target model is also updated adaptively by the principled Bayesian posterior inference, which admits a closed form within our Dirichlet prior modeling. With empirical evaluations on some video datasets, the proposed method is shown to yield more accurate tracking than baseline histogram-based trackers with the same types of features, often being superior to the appearance-based (visual) trackers.

• 산업공학
• /
• 제18권4호
• /
• pp.504-511
• /
• 2005

The NP-hard channel assignment problem becomes more and more important to use channels as efficiently as possible because there is a rapidly growing demand and the number of usable channel is very limited. The hybrid genetic and local search (HGLS) method in this paper is a hybrid method of genetic algorithm with no interference channel assignment (NICA) in clustering stage for diversified search and local search in tuning stage when the step of search is near convergence for minimizing blocking calls. The new representation of solution is also proposed for effective search and computation for channel assignment.

• Interaction and multiscale mechanics
• /
• 제1권1호
• /
• pp.143-156
• /
• 2008

Elliptical tubes may buckle in an elastic local buckling failure mode under uniform compression. Previous analyses of the local buckling of these members have assumed that the cross-section is hollow, but it is well-known that the local buckling capacity of thin-walled closed sections may be increased by filling them with a rigid medium such as concrete. In many applications, the medium many not necessarily be rigid, and the infill can be considered to be an elastic material which interacts with the buckling of the elliptical tube that surrounds it. This paper uses an energy-based technique to model the buckling of a thin-walled elliptical tube containing an elastic infill, which elucidates the physics of the buckling phenomenon from an engineering mechanics basis, in deference to a less generic finite element approach to the buckling problem. It makes use of the observation that the local buckling in an elliptical tube is localised with respect to the contour of the ellipse in its cross-section, with the localisation being at the region of lowest curvature. The formulation in the paper is algebraic and it leads to solutions that can be determined by implementing simple numerical solution techniques. A further extension of this formulation to a stiffness approach with multiple degrees of buckling freedom is described, and it is shown that using the simple one degree of freedom representation is sufficiently accurate for determining the elastic local buckling coefficient.

• Earthquakes and Structures
• /
• 제11권2호
• /
• pp.217-243
• /
• 2016

The linear and nonlinear seismic responses of steel buildings with perimeter moment resisting frames and welded connections (WC) are estimated and compared with those of buildings with post-tensioned connections (PC). Two-dimensional (2D) and three-dimensional (3D) structural representations of the buildings as well as global and local response parameters are considered. The seismic responses and structural damage of steel buildings with PC may be significantly smaller than those of the buildings with typical WC. The reasons for this are that the PC buildings dissipate more hysteretic energy and attract smaller inertia forces. The response reduction is larger for global than for local response parameters. The reduction may significantly vary from one structural representation to another. One of the main reasons for this is that the energy dissipation characteristics are quite different for the 2D and 3D models. In addition, in the case of the 3D models, the contribution of each horizontal component to the axial load on an specific column may be in phase each other during some intervals of time, but for some others they may be out of phase. It is not possible to observe this effect on the 2D structural formulation. The implication of this is that 3D structural representation should be used while estimating the effect of the PC on the structural response. Thus, steel frames with post-tensioned bolted connections are a viable option in high seismicity areas due to the fact that brittle failure is prevented and also because of their reduced response and self-centering capacity.

• 한국자동차공학회논문집
• /
• 제1권1호
• /
• pp.131-139
• /
• 1993

Low cycle fatigue characteristics of cast aluminum alloy A356 with a yield strength and ultimate strength of 229 and 283 MPa respectively was evaluated using smooth axial specimen under strain controlled condition. Reversals to failure ranged from 16 to 107. The cast aluminum alloy exhibited cyclically strain-gardening behavior. The results of low cycle fatigue tests indicated that the conventional low cycle fatigue tests indicated that the conventional low cycle fatigue life model was not a satisfactory representation of the data. This occurred because the elastic strain-life curve was not-log-log linear and this phenomena caused a nonconservative and unsafe fatigue life prediction at both extremes of long and short lives. A linear log-log total strain-life model and a bilinear log-log elastic strain-life model were proposed in order to improve the representation of data compared to the conventional low cycle fatigue life model. Both proposed fatigue life models were statistically analyzed using F tests and successfully satisfied. However, the low cycle fatigue life model generated by the bilinear log-log elastic strain-life equation yielded a discontinuous curve with nonconservatism in the region of discontinuity. Among the models examined, the linear log-log total strain-life model provided the best representation of the low cycle fatigue data. Low cycle fatigue life prediction method based on the local strain approach could conveniently incorporated both proposed fatigue life models.

• 한국정보통신학회논문지
• /
• 제23권6호
• /
• pp.682-689
• /
• 2019

부분 기반 영상 표현(part-based image representation)에서는 영상의 부분적인 모습을 기저 벡터로 표현하고 기저 벡터의 선형 조합으로 영상을 분해하며, 이 때 기저 벡터의 계수가 곧 물체의 부분적인 특징을 의미하게 된다. 본 논문에는 부분 기반 영상 표현 기법인 비음수 행렬 분해(non-negative matrix factorization, NMF)를 이용하여 얼굴 영상을 표현하고 신경망 기법을 적용하여 가려진 얼굴을 인식하는 얼굴 인식을 제안한다. 표준 비음수 행렬 분해, 투영 경사 비음수 행렬 분해, 직교 비음수 행렬 분해를 이용하여 얼굴 영상을 표현하였고, 각 기법의 성능을 비교하였다. 인식기로는 학습벡터양자화 신경망을 사용하였으며, 인식기에서의 거리 척도로는 유클리디언 거리를 사용하였다. 실험 결과, 전통적인 얼굴 인식 방법에 비하여 제안한 기법이 가려진 얼굴 인식에 보다 강인함을 보인다.

• 정보처리학회논문지A
• /
• 제11A권7호
• /
• pp.555-562
• /
• 2004

뇌의 하부 구조인 해마의 전역적 부피 감소와 국부적 형상 변화는 정신의학적 질환에 깊게 관련되어 있다. 해마 구조에 관한 형상 분석 연구는 크게 해마 형상 표현 모델을 구축하고, 이러한 형상 표현으로부터 형상 유사성을 계산하는 과정으로 구성된다. 본 논문에서는 메쉬, 복셀, 골격 데이터를 포함하는 복합적인 옥트리 기반의 형상 표현을 이용하여 해마의 형상을 분석하기 위한 새로운 방법을 제시한다. 우선 해마에 관한 MRI 데이터를 입력으로 받아, 마칭큐브 알고리즘을 사용하여 다해상도 메쉬 모델을 구축한다. 이렇게 구성된 다각형 모델은 깊이맵 기반의 복셀화 방법을 이용하여 중간 단계의 이진 복셀 데이터로 변환된다. 그리고 변환된 복셀 데이터로부터 슬라이스 기반의 골격화 방법에 의하여 해마의 3차원 골격을 추출한다. 그런 후에 옥트리 기반의 다해상도 형상 표현을 얻기위해 해마의 메쉬, 복셀, 골격 데이터를 계층적으로 공간 분할하여 저장하고, 광선 추적 기반의 메쉬 샘플링 방법을 적용하여 샘플 메쉬 데이터를 추출한다. 최종적으로, 형상간 유사성 측정을 위하여 추출된 골격으로부터 방사되는 광선들과 충돌되는 각 샘플 메쉬 쌍에 대하여 $L_2$과 하우스도르프 거리를 계산하고 인터랙티브한 국부적 형상 분석을 지원하기 위하여 마우스 피킹 인터페이스를 채택한다. 이것은 형상의 국부적 변화에 대하여 다양한 해상도에 기반한 형상 분석을 가능하게 한다. 본 논문에서는 실험을 통하여, 제시한 형상 분석 방법이 회전과 스케일 등의 변환에 강인하고, 특히 형상의 국부적 변화 정도를 정확도를 유지하면서 빠르게 평가하는데에 효과적임을 확인하였다. 경로의 수신 신호가 완전 동기 된 수신 신호임을 확인하였다.omonas aeruginosa PA01과 $82\%$로 가장 높은 유사성을 보였고 Pseudomonas arvilla C-1와는 $71\%,$ Pseudomonas putida KT2440과는 $59\%,$ 그리고 Pseudomonas sp. CA10과는 $53\%$의 상동성이 각각 존재하는 것으로 확인하였다.)을 가지고 있음이 확인되었다. 사람에 직접적인 유해성을 가지고 있는 지 확인하기 위해 사람 방광 유래의 T-24세포와 장내 표피 유래의 Caco-2세포에 대한 부착능을 시험하였을 때, 16균주$(42.1\%)$가 T-24방광 세포에, 그리고 17균주$(44.7\%)$가 Caco-2장세포에 대해 강한 부착능을 나타내었다. 특히 11균주$(28.9\%)$는 두 세포 모두에 강한 부착능을 가지고 있었다. Filter mating method를 수행하여 이들 균주들의 독소 생산 유전자와 항생제 내성 유전자가 사람에서 분리된 균주로 전달되는 것을 확인할 수 있었다. 본 실험의 결과는 설사 중상을 나타내는 돼지로부터 분리된 용혈성 E. coli의 독성과 세포 부착능력, 그리고 항생제 내성간의 상호 연관성을 보여주지 않았으나 동물 분리 세균의 항생제 내성과 독소 생산 능력이 유전자 전달을 통해서 뿐만 아니라 세균의 직접 접촉에 의해서도 인체로 전달될 수 있는 것을 보여주는 것이다.다. 본 연구를 토대로 장시간의 체외순환에서는 신장기능을 대표하는 수치들에도

• 대한수학회보
• /
• 제59권4호
• /
• pp.917-928
• /
• 2022

Let R be a commutative Noetherian ring and I an ideal of R. In this paper, we study colocalization of generalized local homology modules. We intend to establish a dual case of local-global principle for the finiteness of generalized local cohomology modules. Let M be a finitely generated R-module and N a representable R-module. We introduce the notions of the representation dimension r^I(M, N) and artinianness dimension a^I(M, N) of M, N with respect to I by r^I(M, N) = inf{i ∈ ℕ₀ : H^I_i(M, N) is not representable} and a^I(M, N) = inf{i ∈ ℕ₀ : H^I_i(M, N) is not artinian} and we show that a^I(M, N) = r^I(M, N) = inf{r^IR_𝔭 (M_𝔭,𝔭N) : 𝔭 ∈ Spec(R)} ≥ inf{a^IR_𝔭 (M_𝔭,𝔭N) : 𝔭 ∈ Spec(R)}. Also, in the case where R is semi-local and N a semi discrete linearly compact R-module such that N/∩_t>0I^tN is artinian we prove that inf{i : H^I_i(M, N) is not minimax}=inf{r^IR_𝔭 (M_𝔭,𝔭N) : 𝔭 ∈ Spec(R)＼Max(R)}.

• 음성과학
• /
• 제10권2호
• /
• pp.287-301
• /
• 2003

The purpose of this study was to examine whether the same vowel sounds are perceived differently by the two local dialect speakers, Seoul dialect speakers (SDS) and Kyungnam dialect speakers (KDS), whose vowel systems differ each other. In the first experiment SDS and KDS heard vowels synthesized in vowel space with F1 by F2 and categorized them into one of 10 Korean monophthongs. The results showed that SDS and KDS perceived the synthesized vowels differently. For example, /$\varepsilon$ versus /e/ contrast, ${\o}$/, and /y/ are differentiated by SDS, whereas they are perceptually confused by KDS. We also observed that /i/ could not be perceived unless the vowel synthesis included F3 or higher formants. In the second experiment SDS and KDS performed the similarity rating task of 10 synthesized Korean monophthongs. Two-dimensional MDS solution based on the similarity rating scores was obtained for each dialect group. The first dimension can be named 'vowel advancement' and the second 'vowel height'. The comparison of the two MDS solutions showed that the overall psychological distances among the vowels are shorter in KDS than SDS and that especially the distance between /$\Lambda$/ and /i/ is shorter in KDS than SDS. The result suggested that perception or mental representation of vowels depends on the vowel system of the listener's dialect or language. Further research problems were discussed in the final section.