OpenCV findChessboardCorners very slow

Question

dose anybody know why is cv::findChessboardCorners horribly slow on all images with more than one mega-pixel ( > 1 000 000 pixles ).

As a workaround I down-sample my images to fit this bound, but I would like to know if it can be solved more "elegantly".

Thanx

score 1 · Answer 1 · edited Dec 02 '19 at 14:10

1

Haven't dug into the reasons - I wrote a much faster square finder from scratch.

The circleGrid provides a better result and is much faster, it uses the new BlobDetector interface.

edited Dec 02 '19 at 14:10

Wolfgang Fahl

answered Nov 21 '12 at 18:10

Martin Beckett

it's in the examples dir in opencv - stereocalib.cpp (IIRC) – Martin Beckett Nov 23 '12 at 16:03

1 Answers1