• Title, Summary, Keyword: 정역학 관계

Search Result 7, Processing Time 0.082 seconds

Face Detection And Recognition using Hidden Markov Models (HMM 을 이용한 얼굴 검출과 인식)

  • 박호석;차영석;최현수;배철수;권오홍;최철재;나상동
    • Proceedings of the Korean Institute of Information and Commucation Sciences Conference
    • /
    • /
    • pp.336-341
    • /
    • 2000
  • Hidden Markov Model(HMM)을 기반으로 한 얼굴 검출과 얼굴 인식에 대한 프레임작업에 대한 것이다. 관찰 벡터는 Karhunen-Loves Transform(KLT)의 상관관계를 이용하여 얻은 HMM의 정역학 특성을 사용하였으며, 본 연구에서 보여준 얼굴인식 방법은 이전의 HMM 기반의 얼굴인식 방법에서 인식률을 약간 개선함으로써 컴퓨터 연산을 훨씬 간단히 할 수 있음을 보여준다

  • PDF

極小 Energy 定理와 그 應용 (I)

  • 양원호
    • Journal of the KSME
    • /
    • v.20 no.3
    • /
    • pp.211-217
    • /
    • 1980
  • 탄성학문제의 엄밀해는 응력의 평형방정식과 변형의 적합조건식 또는 이들을 조합한 탄성의 기 초방정식을 만족하며, 주어진 경계조건을 만족하는 해를 구해야 하겠지만, 문제에 따라서는 그 엄밀해를 구하기가 곤란하거나 또는 아주 복잡하므로 엄밀해에 가까운 근사해를 구하는 것이 편리할 때가 있다. 본강좌에서는 극소 energy 정리와 ritz의 근사계산법을 결합하여 탄성문제의 근사해를 구하는 방법을 설명하고자 한다. 강좌의 처음에는 삼차원에서의 변형 energy와 외력의 일(work)을 유도하고, 이들 사이의 관계로부터 일반국소 energy 정리를 정의한 다음 이 정리가 실제문제에 어떻게 응용될 수 있는가를 보이는 응용예를 주로하여 진행해 보려한다. 이때의 응 용예로 서는 재료역학에서 이미 눈에 익은 기초적 문제를 주로 다루어 보려한다. 재료역학에서의 탄성문제의 해는 정정인 문제와 불정정인 문제를 따로 분류하며, 불정정인 문제의 해는 정역학의 평형방정식과 변형의 적합방정식을 연립으로 하여 해결하든가, 중첩법을 적용하므로서 일반적 으로 상당히 복잡한 해가 되는 것이 보통인데, 본강좌에서 기술하는 방법은 정정 불정정의 문 제를 구별할 필요가 없이 같은 방법이 적용되며 어떤 면에서는 불정정의 문제가 정정의 문제보다 그 해가 간편히 구해질 수 있다는 장점이 있는 것이다.

  • PDF

A Face Recognition using the Hidden Markov Model and Karhuman Loevs Transform (Hidden Markov Model과 Karhuman Loevs Transform를 이용한 얼굴인식)

  • Kim, Do-Hyun;Hwang, Suen-Ki;Kang, Yong-Seok;Kim, Tae-Woo;Kim, Moon-Hwan;Bae, Cheol-Soo
    • The Journal of Korea Institute of Information, Electronics, and Communication Technology
    • /
    • v.4 no.1
    • /
    • pp.3-8
    • /
    • 2011
  • The work presented in this paper describes a Hidden Markov Model(HMM)-based framework for face recognition and face detection. The observation vectors used to characterize the statics of the HMM are obtained using the coefficients of the Karhuman-Loves Transform(KLT). The face recognition method presented in this paper reduces significantly the computational complexity of previous HMM-based face recognition systems, while slightly improving the recognition rate. In addition, the suggested method is more effective than the exiting ones in face extraction in terms of accuracy and others even under complex changes to the surroundings such as lighting.

The Forward/Inverse Force Transmission Analyses of the Stewart Platform (스튜어트 플랫폼의 순방향/역방향 힘 전달 해석)

  • Kim, Han-Sung;Choi, Yong-Je
    • Journal of the Korean Society for Precision Engineering
    • /
    • v.16 no.5
    • /
    • pp.200-208
    • /
    • 1999
  • The statics relation of the Stewart platform has been investigated from the viewpoint of the forward and inverse force transmission analyses. Two eigenvalue problems corresponding to the forward and inverse force transmission analyses have been formulated. The forward force transmission analysis is to determine the ranges of the magnitudes of the force and moment generated at the end-effector for the given magnitude of linear actuator forces. In reverse order, the inverse force transmission analysis is to find the range of the magnitude of actuator forces for the given ranges of the magnitudes of the force and moment at the end-effector. The inverse force transmission analysis is important since it can provide a designer with a valuable information about how to choose the linear actuators. It has been proved that two eigenvalue problems have a reciprocal relation, which implies that solving either of the eigenvalue problems may complete the forward/inverse force transmission analysis. A numerical example has been also presented.

  • PDF

Stiffness Analysis of a Low-DOF Parallel Manipulator using the Theory of Reciprocal Screws (역나선 이론을 이용한 저자유도 평행구조 기구의 강성해석)

  • Kim, Han-Sung
    • Proceedings of the KSME Conference
    • /
    • /
    • pp.573-578
    • /
    • 2004
  • This paper presents a methodology for the stiffness analysis of a low-DOF parallel manipulator. A low-DOF parallel manipulator is a spatial parallel manipulator which has less than six degrees of freedom. The reciprocal screws of actuations and constraints in each leg can be determined by making use of the theory of reciprocal screws, which provide information about reaction forces due to actuations and constraints. When pure force is applied to a leg, the leg stiffness is modeled as a linear spring along the line. For pure couple, it is modeled as a rotational spring about the axis. It is shown that the stiffness model of an F-DOF parallel manipulator consists of F springs related to actuations and 6-F springs related to constraints connected from the moving platform to the base in parallel. The $6{\times}6$ Cartesian stiffness matrix is obtained, which is the sum of the Cartesian stiffness matrices of actuations and constraints. Finally, a 3-UPU parallel manipulator is used as an example to demonstrate the methodology.

  • PDF

Face Recognition Method Robust to Change in Lighting Condition (조명의 변화에 강건한 얼굴인식)

  • Nam, Kee-Hwan;Han, Jun-Hee;Park, Ho-Sik;Lee, Young-Sik;Jung, Yen-Gil;Ra, Sang-Dong;Bae, Cheol-Soo
    • Proceedings of the Korean Institute of Information and Commucation Sciences Conference
    • /
    • v.9 no.1
    • /
    • pp.1137-1140
    • /
    • 2005
  • The work presented in this paper describes a Hidden Markov Model(HMM)-based framework for face recognition and face detection. The observation vectors used to characterize the statics of the HMM are obtained using the coefficients of the Karhuman-Loves Transform(KLT). The face recognition method presented in this paper reduces significantly the computational complexity of previous HMM-based face recognition systems, while slightly improving the recognition rate. In addition, the suggested method is more effective than the exiting ones in face extraction in terms of accuracy and others even under complex changes to the surroundings such as lighting.

  • PDF

Stiffness Analysis of a Low-DOE Parallel Manipulator using the Theory of Reciprocal Screws (역나선 이론을 이용한 저자유도 병렬형 기구의 강성해석)

  • Kim Han Sung
    • Transactions of the Korean Society of Mechanical Engineers A
    • /
    • v.29 no.5
    • /
    • pp.680-688
    • /
    • 2005
  • This paper presents a methodology for the stiffness analysis of a low-DOF parallel manipulator. A low-DOF parallel manipulator is a spatial parallel manipulator which has less than six degrees of freedom. The reciprocal screws of actuations and constraints in each leg can be determined by making use of the theory of reciprocal screws, which provide information about reaction forces due to actuations and constraints. When pure farce is applied to a leg, the leg stiffness is modeled as a linear spring along the line. For pure couple, it is modeled as a rotational spring about the axis. It is shown that the stiffness model of an it_DOF parallel nipulator consists of F springs related to actuations and 6-F springs related to constraints connected from the moving platform to the base in parallel. The 6x f Cartesian stiffness matrix is derived, which is the sum of the Cartesian stiffness matrices of actuations and constraints. Finally, the 3-UPU, 3-PRRR, and Tricept parallel manipulators are used as examples to demonstrate the methodology.