Ask Your Question

Calibrating a camera with ChessboardCorners

asked 2013-06-04 23:50:31 -0500

eyal gravatar image

Hi All,

I've read in the book "Learning OpenCV" by O'REILLY press and from "OpenCV 2 Computer Vision Application Programming Cookbook" by Robert Laganière (PACKT publishing) that when calibrating camera from chessboard pattern they use for 3D world coordinates the cornder indices with z = 0.

for example suppose chesboard cornders in vertical dimension is 4 and horizontal dimension 5 so the corresponding 3D world coordinates for calibration become:

(0,0,0), (0,1,0), (0,2,0),(0,3,0), (0,4,0), (1,0,0),(1,1,0),(1,2,0)...and so on. I understand we we can pick Z= 0 but I dont understand the one unit for each cornder coordinates I mean what if each sqaure in my image is say for example 44 cm? How does it reflects in the above calibration process?

Thank u in advance

edit retag flag offensive close merge delete


Why are you caring of the square size? Please, clarify your question.

Daniil Osokin gravatar imageDaniil Osokin ( 2013-06-05 02:11:41 -0500 )edit

Do you want to know measurement unit for points coordinates?

Daniil Osokin gravatar imageDaniil Osokin ( 2013-06-05 03:08:49 -0500 )edit

I don't see how the actual square size(real world size) is taken into account in the calibration process when using (i,j,0) for cornder at row i and column j.

eyal gravatar imageeyal ( 2013-06-05 15:18:24 -0500 )edit

1 answer

Sort by » oldest newest most voted

answered 2013-06-06 00:43:52 -0500

Daniil Osokin gravatar image

Hi, to generate ideal pattern corners you need the square size, then algorithm will find correspondence between ideal pattern corners and distorted (captured by camera) and estimate intrinsic parameters. Try the Camera calibration tutorial, it uses square size.

edit flag offensive delete link more
Login/Signup to Answer

Question Tools


Asked: 2013-06-04 23:50:31 -0500

Seen: 211 times

Last updated: Jun 06 '13