본문으로 바로가기

공돌이 연습장

현재위치 :: HOME BLOG CATEGORY SEARCH ARCHIVE TAGS MEDIA LOCATION GUESTBOOK

네비게이션

  • 홈
  • 태그
  • 방명록
관리자
  • 블로그 이미지
    shicnova

    IT 관련 관심사 및 공부내용 정리를 위한 티스토리입니다.

    링크추가
  • 글쓰기
  • 환경설정
  • 로그인
  • 로그아웃

HSV Color model (RGB to HSV)

HSV cylinder Range of Hue : [0 - 2π)Range of Saturation : [0 - 1)Range of Value : [0 - 255) Hue and chromaFig. 9. Both hue and chroma are defined based on the projection of the RGB cube onto a hexagon in the "chromaticity plane". Chroma is the relative size of the hexagon passing through a point, and hue is how far around that hexagon’s edge the point lies.In each of our models, we calculate bot..

2016-1/Image Processing 2016. 10. 2. 06:32

System model & Belief distribution

System Models 일반적인 linear control system 모델은 state transition 방정식과 observation 방정식을 갖는데, 다음과같다. 확률 선형모델에서는 노이즈가 추가되는데, 다음과 같은 방정식이 있다. State Transition 이러한 상태변환식의 Complete State 형태는 로 나타내어진다. Observation system에 있어 state x_t는 직접적으로 관측될 수 없어서 observation( i.e. measurement) z_t를 사용한다. 이 관측의 Complete State 형태는 로 나타내어진다. 확률에서 State Transition 과 Observation을 나타내어보면 마찬가지로 Specific measurement의 경우 두 경우..

2016-2/Robotics & Intelligent System 2016. 9. 24. 06:30

Marcov Process

Marcov Process {x0, x1, … , xt-1, xt, xt+1, … } 를 랜덤변수 X의 sample 또는 status의 sequence라 하자. time t 에서의 상태 probability가 오직 t-1 일때의 상태에만 종속(dependent)된다면 이 과정을 Marcov Process라 한다. 식으로 표현하면, Marcov Process의 대표적인 예로는 Wiener process or Brownian motion 이 있다. 또한 반복적인 recursive estimation 에도 쓰이는데, 대표적으로 Bayes filter – a Kalman filter or a particle filter 에 적용될 수 있다. Marcov chain

2016-2/Robotics & Intelligent System 2016. 9. 24. 05:10

Maximum Likelihood Estimator

최대 우도 추정법 (Maximum Likelihood Estimator) Random Sample X1, X2,..., Xn 를 갖고있다고 가정.또한 probability distribution 은 θ 라는 파라미터에 의해서 결정된다고 하자. 정의에 의해, Likelihood Function L(θ)는 다음과 같이 표현된다. 연산에 의해 정리하면, 발생 가능성 이라고도 하는데, 이 함수의 최대치를 찾기 위해, 위 식을 미분한다. 표본의 수가 많고, 이에따라 정규분포의 형태를 이룬다고 할 때, 양 변을 미분하여 0인 값을 갖는 θ값을 구한다. 이러한 θ값이 Maximum Likelihood Estimator가 된다.

2016-2/Robotics & Intelligent System 2016. 9. 24. 04:51

Central Limit Theorem (CLT)

For any distribution, 모집단의 분산이 어떠한 형태를 나타내던, (Uniform, Gaussian, etc.) 표본의 수가 증가하면 할수록 표본의 분포는 정규분포를 따르게 된다. N(μ,σ 2) = Gaussian Distribution (Image referecne : https://en.wikipedia.org/wiki/Central_limit_theorem)

2016-2/Robotics & Intelligent System 2016. 9. 24. 04:30
  • 이전
  • 1
  • 2
  • 3
  • 다음

사이드바

NOTICE

  • 전체 보기
MORE+

CATEGORY

  • 분류 전체보기 (12)
    • IT 이야기 (2)
      • News (1)
      • Issue (1)
    • 2016-1 (3)
      • TCP IP Kernel Seminar (0)
      • Routing Seminar (0)
      • Image Processing (3)
    • Paper Review (0)
    • 2016-2 (7)
      • Linear Algebra (2)
      • Vector & Matrices (0)
      • Random Proecssing (0)
      • Imge Processing (0)
      • Robotics & Intelligent Syst.. (5)

RECENTLY

  • 최근 글
  • 최근 댓글

최근 글

최근댓글

Trackback

TAG

MORE+

ARCHIVE

CALENDAR

«   2025/05   »
일 월 화 수 목 금 토
1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31

LINK

VISITOR

오늘
어제
전체
  • 홈으로
  • 방명록
  • 로그인
  • 로그아웃
  • 맨위로
SKIN BY COPYCATZ COPYRIGHT 공돌이 연습장, ALL RIGHT RESERVED.
공돌이 연습장
블로그 이미지 shicnova 님의 블로그
MENU
  • 홈
  • 태그
  • 방명록
CATEGORY
  • 분류 전체보기 (12)
    • IT 이야기 (2)
      • News (1)
      • Issue (1)
    • 2016-1 (3)
      • TCP IP Kernel Seminar (0)
      • Routing Seminar (0)
      • Image Processing (3)
    • Paper Review (0)
    • 2016-2 (7)
      • Linear Algebra (2)
      • Vector & Matrices (0)
      • Random Proecssing (0)
      • Imge Processing (0)
      • Robotics & Intelligent Syst.. (5)
VISITOR 오늘 / 전체
  • 글쓰기
  • 환경설정
  • 로그인
  • 로그아웃
  • 취소

검색

티스토리툴바