Notice
Recent Posts
Recent Comments
Link
| 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 |
Tags
- rostopics
- imageprocessing
- ComputerVision
- Slam
- roslaunch
- Ros
- gaussiansplatting
- LIDAR
- tilebasedrasterizer
- rosnoetic
- vectorfields
- turtlebot3
- adaptivedensitycontrol
- vectorcalculus
- turtlebot
- raspberrypi
- 3dgaussiansplatting
- catkinworkspace
- sensorfusion
- usbcamera
- 3dmapping
- NERF
- opencv
- electromagnetism
- realtimerendering
- differentiablerendering
- alphablending
- pointcloud
- rospackages
- covariancematrix
Archives
- Today
- Total
Wiredwisdom
`A=SLamdaS^-1, B=QLamdaQ^T` 본문
The eigenvectors of Symmetric matrix are Perpendicular
General Case
`A=SLamdaS^-1`
`A=` Random Square Matrix
`S=` EigenVector-Matrix, not Orthornomal!
`Lamda=` EigenValue Matrix (Real Number)
Symmetric Matrix Case
`B=QLamdaQ^-1=QLamdaQ^T`
`B=` Symmetric Matrix
`Q=` EigenVector, Orthonormal Matrix(Symmetric Matrix)
`Lamda=` EigenValue(Real Number)
When Symmetric Matrix
`Q^-1=Q^T`
그래서 Rotation Matrix 는 대칭 행렬인지도 모른다.
그리고 공분산의 경우, 여기서의 EigenVector는 정보량의 밀집정도를 나타내는데
3개의 벡터의 경우 데이터 밀도가 가장 높은 축 2개, 그리고 데이터 벡터의 밀도가 가장 낮은 축을 나머지로 하여
Value 값으로 이를 구분할 수 있다.
Value 값이 가장 낮으면 해당 EigenVector에는 데이터가 가장 적으므로,
이를 PointCloud 에서는 normal 값, 즉 표면에 수직한 법선으로 표기한다.
'MATH > Linear Algebra' 카테고리의 다른 글
| `ABA^T` 형식의 활용 (0) | 2024.06.24 |
|---|---|
| The geometry of eigenvalues and eigenvectors (0) | 2024.01.20 |
| Geometry of the Inverse Matrix (0) | 2024.01.20 |