3

I am using python 3 and opencv3.6 and i am performing a stereo camera calibration using a chessboard.

I was asking me, there is a moment where i have to give the square size? I checked the opencv tutorial but it wasn't very clear. I would like a confirm.

If the criteria i am using is : 

criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 30, 0.001)

The 30 stays for the square size??

The function i am using are cv2.calibrateCamera, cv2.stereoCalibrate

 ret, mtx, dist, mtx2, dist2, R, T, E, F = cv2.stereoCalibrate(objpoints_l, imgpoints_l ,imgpoints_r, mtxL, distL, mtxR, distR,
                                                          gray_l.shape[::-1],
                                                          flags = flags,
                                                          criteria = termination_criteria_extrinsics )

Thanks and sorry for the noob question

gaetano ma
  • 139
  • 12
  • Please read the [doc](https://docs.opencv.org/master/d9/d0c/group__calib3d.html#ga91018d80e2a93ade37539f01e6f07de5). 30 is the maximum number of iteration. You should specify the coordinates of your square chessboard in `objpoints`. See the [example](https://github.com/opencv/opencv/blob/master/samples/cpp/stereo_calib.cpp) – Miki May 31 '18 at 15:13
  • so I give just the number of raw and column of the squares but not their dimension, right? – gaetano ma May 31 '18 at 15:30
  • @Miki I can't find anything like "objpoints" in the docs. Please clarify – Eshita Shukla Jul 28 '18 at 05:29

0 Answers0