• 대한전자공학회논문지SP
• /
• 제43권2호
• /
• pp.82-86
• /
• 2006

패턴인식 시스템은 일반적으로 데이터의 전처리, 특징 추출, 학습단계의 과정을 거쳐서 개발되어 진다. 그중에서도 특징 추출 과정은 다차원 공간을 가진 입력 데이터의 복잡도를 줄여서 다음 단계인 학습단계에서 계산 복잡도와 인식률을 향상시키는 역할을 한다. 패턴인식에서 특징 추출 기법으로써 principal component analysis, factor analysis, linear discriminant analysis 같은 방법들이 널리 사용되어져 왔다. 이 논문에서는 singular value decomposition (SVD) 방법이 패턴인식 시스템의 특징 추출과정에 유용하게 사용될 수 있음을 보인다. 특징 추출단계에서 SVD 기법의 유용성을 검증하기 위하여 원격탐사 응용에 적용하였는데, 실험결과는 널리 쓰이는 PCA에 비해 약 25%의 인식률의 향상을 가져온다는 것을 알 수 있다.

• Journal of electromagnetic engineering and science
• /
• 제5권2호
• /
• pp.49-60
• /
• 2005

In this paper, an efficient performance enhancement scheme for an adaptive antenna array under the flat and the frequency-selective Rayleigh fadings is proposed. The proposed signal enhancement scheme is the modified linear signal estimator which combines the rank N approximation by reducing noise eigenvalues(RANE) and Toeplitz matrix approximation(TMA) methods into the linear signal estimator. The proposed performance enhancement scheme is performed by not only reducing the noise component from the signal-plus-noise subspace using RANE but also having the theoretical property of noise-free signal using TMA. Consequently, the key idea of the proposed performance enhancement scheme is to greatly enhance the performance of an adaptive antenna array by removing all undesired noise effects from the post-correlation received signal. The proposed performance enhancement scheme applies at the Wiener maximal ratio combining(MRC) method which has been widely used as the conventional adaptive antenna array. It is shown through several simulation results that the performance of an adaptive antenna array using the proposed signal enhancement scheme is much superior to that of a system using the conventional method under several environments, i.e., a flat Rayleigh fading, a fast frequency-selective Rayleigh fading, a perfect/imperfect power control, a single cell, and a multicell.

• 대한수학회보
• /
• 제23권2호
• /
• pp.91-102
• /
• 1986

A topological space (X,.tau.) is called invertible [7] if for each proper open set U in (X,.tau.) there exists a homoemorphsim h:(X,.tau.).rarw.(X,.tau.) such that h(X-U).contnd.U. Doyle and Hocking [7] and Levine [13], as well as others have investigated properties of invertible spaces. Recently, Crosseley and Hildebrand [5] have introduced the concept of semi-invertibility, which is weaker than that of invertibility, by replacing "homemorphism" in the definition of invertibility with "semihomemorphism", A space (X,.tau.) is said to be semi-invertible if for each proper semi-open set U in (X,.tau.) there exists a semihomemorphism h:(X,.tau.).rarw.(X,.tau.) such that h(X-U).contnd.U. The purpose of the present article is to introduce the class of almost-invertible spaces containing the class of semi-invertible spaces and to investigate its properties. One of the primary concerns will be to determine when a given local property in an almost-invertible space is also a global property. We point out that many of the results obtained can be applied in the cases of semi-invertible spaces and invertible spaces. For example, it is shown that if an invertible space (X,.tau.) has a nonempty open subset U which is, as a subspace, H-closed (resp. lightly compact, pseudocompact, S-closed, Urysohn, Urysohn-closed, extremally disconnected), then so is (X,.tau.).hen so is (X,.tau.).

• Journal of Electrical Engineering and Technology
• /
• 제13권6호
• /
• pp.2511-2520
• /
• 2018

In this study, Multi-Radial Basis Function Support Vector Machine (Multi-RBF SVM) classifier is introduced based on a composite kernel function. In the proposed multi-RBF support vector machine classifier, the input space is divided into several local subsets considered for extremely nonlinear classification tasks. Each local subset is expressed as nonlinear classification subspace and mapped into feature space by using kernel function. The composite kernel function employs the dual RBF structure. By capturing the nonlinear distribution knowledge of local subsets, the training data is mapped into higher feature space, then Multi-SVM classifier is realized by using the composite kernel function through optimization procedure similar to conventional SVM classifier. The original training data set is partitioned by using some unsupervised learning methods such as clustering methods. In this study, three types of clustering method are considered such as Affinity propagation (AP), Hard C-Mean (HCM) and Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA). Experimental results on benchmark machine learning datasets show that the proposed method improves the classification performance efficiently.

• Smart Structures and Systems
• /
• 제17권3호
• /
• pp.523-539
• /
• 2016

The ambient vibration measurement is an output-data-only dynamic testing where natural excitations are represented, for instance, by winds and typhoons. The modal identification involving output-only measurements requires the use of specific modal identification techniques. This paper presents the application of a reliable method (the Stochastic Subspace Identification - SSI) implemented in a general purpose software. As a criterion toward the robustness of identified modes, a bio-inspired optimization algorithm, with a highly nonlinear objective function, is introduced in order to find the optimal deployment of a reduced number of sensors across a large civil engineering structure for the validation of its modal identification. The Ting Kau Bridge (TKB), one of the longest cable-stayed bridges situated in Hong Kong, is chosen as a case study. The results show that the proposed method catches eigenvalues and eigenvectors even for a reduced number of sensors, without any significant loss of accuracy.

• Smart Structures and Systems
• /
• 제10권4_5호
• /
• pp.353-373
• /
• 2012

The instrumented Canton Tower is a 610 m high-rise structure, which has been considered as a benchmark problem for structural health monitoring (SHM) research. In this paper, an improved automatic modal identification method is presented based on a natural excitation technique in conjunction with the eigensystem realization algorithm (NExT/ERA). In the proposed modal identification method, damping ratio, consistent mode indicator from observability matrices (CMI_O) and modal amplitude coherence (MAC) are used as criteria to distinguish the physically true modes from spurious modes. Enhanced frequency domain decomposition (EFDD), the data-driven stochastic subspace identification method (SSI-DATA) and the proposed method are respectively applied to extract the modal parameters of the Canton Tower under different environmental conditions. Results of modal parameter identification based on output-only measurements are presented and discussed. User-selected parameters used in those methods are suggested and discussed. Furthermore, the effect of environmental conditions on the dynamic characteristics of Canton tower is investigated.

• 대한수학회지
• /
• 제48권2호
• /
• pp.397-420
• /
• 2011

In this paper we study the structure of closed weakly dense ideals in Privalov spaces $N^p$ (1 < p < $\infty$) of holomorphic functions on the disk $\mathbb{D}$ : |z| < 1. The space $N^p$ with the topology given by Stoll's metric [21] becomes an F-algebra. N. Mochizuki [16] proved that a closed ideal in $N^p$ is a principal ideal generated by an inner function. Consequently, a closed subspace E of $N^p$ is invariant under multiplication by z if and only if it has the form $IN^p$ for some inner function I. We prove that if $\cal{M}$ is a closed ideal in $N^p$ that is dense in the weak topology of $N^p$, then $\cal{M}$ is generated by a singular inner function. On the other hand, if $S_{\mu}$ is a singular inner function whose associated singular measure $\mu$ has the modulus of continuity $O(t^{(p-1)/p})$, then we prove that the ideal $S_{\mu}N^p$ is weakly dense in $N^p$. Consequently, for such singular inner function $S_{\mu}$, the quotient space $N^p/S_{\mu}N^p$ is an F-space with trivial dual, and hence $N^p$ does not have the separation property.

• 한국콘텐츠학회논문지
• /
• 제9권2호
• /
• pp.115-124
• /
• 2009

얼굴 인식 방법 중 인기 있는 고유얼굴 기반 얼굴 인식 방법은 훈련 얼굴 이미지 세트에 대해 PCA를 적용하여 얻어진 고유얼굴을 이용한다. 따라서 훈련 얼굴 이미지들의 조명들과 다른 조명의 환경들에서는 신뢰성 있는 성능을 얻기 어렵다. 본 논문에서는 조명의 영향을 배제한 조명분리 고유얼굴 기반 얼굴 인식 방법을 제안한다. 제안된 방법은 얼굴 모델 이미지 세트의 고유얼굴 공간을 구성된 얼굴 조명 부분공간에 대해 직교 분해하여 얻은 조명분리 고유얼굴들을 이용한다. 실험을 통해서 조명분리 고유얼굴에 기반하는 제안된 얼굴 인식 방법이 기존 고유얼굴 기반 얼굴 인식 방법보다 조명의 영향에 보다 강인함을 확인하였다.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제32권8호
• /
• pp.759-768
• /
• 2005

본 논문에서는 다중 혼합 특징 정보를 저 차원 제스처 심볼로 구성하여 제스처를 인식하는 알고리즘에 대해 기술한다. 기존의 기하학적인 특징 기반 방법이나 외관기반 방법에서는 깔, 다리의 위치나 몸의 형상 정보만을 특징 값으로 이용하기 때문에 유사한 신체 동작이나 신체 부위의 움직임에 따라 애매한 결과를 나타내었지만 제안한 방법은 신체의 어느 부위가 움직이는지를 나타내는 부분특징정보(partial feature information)와 전체적인 신체의 형상을 표현하는 전역특징정보(global feature information)를 이용함으로써 동작의 구분뿐만 아니라 유사한 동작을 인식할 수 있는 장점이 있다. 그리고 비교적 적은 계산량과 높은 인식률 때문에 감시 시스템이나 지적 인터페이스 시스템 같은 여러 응용 분야에 적용될 수 있다.

• 대한수학회지
• /
• 제51권5호
• /
• pp.1089-1104
• /
• 2014

An operator $T{\in}L(H)$, is said to belong to k-quasi class $A_n^*$ operator if $$T^{*k}({\mid}T^{n+1}{\mid}^{\frac{2}{n+1}}-{\mid}T^*{\mid}^2)T^k{\geq}O$$ for some positive integer n and some positive integer k. First, we will see some properties of this class of operators and prove Weyl's theorem for algebraically k-quasi class $A_n^*$. Second, we consider the tensor product for k-quasi class $A_n^*$, giving a necessary and sufficient condition for $T{\otimes}S$ to be a k-quasi class $A_n^*$, when T and S are both non-zero operators. Then, the existence of a nontrivial hyperinvariant subspace of k-quasi class $A_n^*$ operator will be shown, and it will also be shown that if X is a Hilbert-Schmidt operator, A and $(B^*)^{-1}$ are k-quasi class $A_n^*$ operators such that AX = XB, then $A^*X=XB^*$. Finally, we will prove the spectrum continuity of this class of operators.