Skip to content
GitLab
  • Menu
Projects Groups Snippets
  • Help
    • Help
    • Support
    • Community forum
    • Submit feedback
    • Contribute to GitLab
  • Sign in / Register
  • L laser_plane_scanner
  • Project information
    • Project information
    • Activity
    • Labels
    • Members
  • Repository
    • Repository
    • Files
    • Commits
    • Branches
    • Tags
    • Contributors
    • Graph
    • Compare
  • Issues 0
    • Issues 0
    • List
    • Boards
    • Service Desk
    • Milestones
  • Merge requests 0
    • Merge requests 0
  • Deployments
    • Deployments
    • Releases
  • Monitor
    • Monitor
    • Incidents
  • Analytics
    • Analytics
    • Value stream
    • Repository
  • Wiki
    • Wiki
  • Snippets
    • Snippets
  • Activity
  • Graph
  • Create a new issue
  • Commits
  • Issue Boards
Collapse sidebar
  • Cmíral Jakub
  • laser_plane_scanner
  • Wiki
  • find_camera_mtx

find_camera_mtx · Changes

Page history
cmirajak deleted page: laser_pos_est authored Oct 19, 2017 by Cmíral Jakub's avatar Cmíral Jakub
Hide whitespace changes
Inline Side-by-side
find_camera_mtx.md
View page @ 9854ef6c
......@@ -10,11 +10,11 @@ We are looking for the Camera Matrix **K** and Distortion Coefficients **D** des
where ![alt text](http://mathurl.com/zbyrpbe.png) are camera focal lengths, ![alt text](http://mathurl.com/jchq2he.png) are optical centers, *p* and *k* are distortion coefficients.
### Calibration object
We chose classic black-white chessboard pattern with 8 rows, 7 columns and square's size approximately 35x35mm (Exact dimensions are specified in source code) as a calibration object.
### Calibration pattern
We chose classic black-white chessboard pattern with eight rows, seven columns and square's size approximately 35x35mm (Exact dimensions are specified in source code) as a calibration object.
### Calibration pictures
We must take pictures with calibration object in the camera's Field of View (FoV) in many different positions to cover most of FoV. Moreover, we must respect that the whole chessboard must be visible.
We must take pictures with calibration object in the camera's Field of View (FoV) in many different positions to cover most of the FoV. Moreover, we must respect that the whole chessboard must be visible.
### Finding chessboards
We use OpenCV function `findChessboarCorners(gray,(rows,column),params=None)->ret,corners` for finding the chessboard in pictures, where `corners` are chessboard's internal corners in pixel (u,v) coordinates. After the chessboard is found, the position of corners is made more precise via OpenCV `cornerSubPixel(gray,corners,winSize,zeroZone,criteria)->None`, which finds them with sub-pixel precision. You can read more about this on the page listed above.
......@@ -22,8 +22,9 @@ We use OpenCV function `findChessboarCorners(gray,(rows,column),params=None)->re
### Camera Matrix and Distortion Coefficients
When we know the position of chessboards in pictures (imagePoints) and correct size of squares and its positions (objectPoints), we can now estimate the Camera Matrix and Distortion Coefficients with OpenCV function `calibrateCamera(objectPoints,imagePoints,imageShape,None,None)->ret,cam_mtx,dist,rvect,tvect`, where `cam_mtx` is the wanted Camera Matrix and `dist` are Distortion coefficients.
### ChAruCo callibration pattern
The calibration using ChAruCo can be initialized using `CameraCalibrationLiveCharuco`. The parameters of a board must be specified inside the script. The package `Calibration.CharucoCameraCalibration` creates the board using `generate_charuco_board(dict, rows, columns, square length, marker length)`, where dict is an OpenCV3 AruCo dictionary, the rest is a shape of board and size of squares in meters.
\ No newline at end of file
Clone repository
  • estimate_laser_plane
  • find_camera_mtx
  • Home
  • laser_trace
  • new_base_coords
  • rotation_table
  • simple setup