Pinhole camera model

Parent: SLAM Index Backlinks: [Camera calibration](camera calibration.md), [pinhole camera projection function](pinhole camera projection function.md), [weiss thesis vision based navigation for micro helicopters](weiss thesis vision-based-navigation-for-micro-helicopters.md) See also: World to camera trafo

Source: http://de.mathworks.com/help/vision/ug/camera-calibration.html Does not account for lens distortion (ideal pinhole camera doesn’t have a lens) To represent a real camera, the full camera model to be used should include (radial and tangential) lens distortion, (such as the one used in the MATLAB computer vision toolbox)

Camera matrix P

• Parameters are represented by a 4x3 camera matrix
• Maps 3Dp -> 2Dp (image plane)
• Extrinsic parameters
  • location of the camera in the 3D world
  • maps 3Dp (world) -> camera coordinates
• Intrinsic parameters
  • non-positional parameters
  • maps camera coords -> 2Dp (image)
  • optical centre (principal point), focal length, skew coefficient

Intrinsic parameters K =