Create repository for camera calibration

[vortexuser]

Initially it should be for Checkerboard, then we can also add support for ChAruco board, and have support for different models.

[vortexuser]

Research which ChAruco board we should purchase

[kluge7]

Idea is to make a camera calibartion repository. You should be able to upload a dataset (images taken by the camera of a checkerboard), and then calculate camera intrinsics and distortion (for example by using OpenCV). It should be modular enough so that you can specify which distortion model you want to use, and which parameters it should and shouldnt estimate.

[kluge7]

Dont use robotsyn due to it being copyright))

[kluge7]

Should have functionality for removing images that have a mean error above X (should be configurable). Images with high mean error can dominate the results you get (see example):

Calibration results

Focal length and principal point

fx: 2335.59214 +/- 24.61281
fy: 2152.01213 +/- 21.33634
cx: 326.21466 +/- 3.00574
cy: 307.67018 +/- 1.87679

Distortion coefficients

k1: 8.204647 +/- 0.171797
k2: -65.136367 +/- 4.221148
p1: 0.243582 +/- 0.007674
p2: -0.283869 +/- 0.010654
k3: 1526.551095 +/- 102.092496

[kluge7]

OpenCV camera calibration examples: https://docs.opencv.org/4.x/dc/dbb/tutorial_py_calibration.html
OpenCV for ChAruco: https://docs.opencv.org/4.x/da/d13/tutorial_aruco_calibration.html