Fisheye calibration image not detecting any corners

943 ●9 ●6

I have the following barrel-distorted fisheye image:

image description

And I am finding the corners in it using the findChessboardCorners function as follows:

cv2.findChessboardCorners(gray, CHECKERBOARD, cv2.CALIB_CB_ADAPTIVE_THRESH | cv2.CALIB_CB_FAST_CHECK | cv2.CALIB_CB_NORMALIZE_IMAGE)

However, I'm getting no corners, i.e., this function returns False, None. Why is that?

-1

berak
32993 ●7 ●81 ●312

your image is not a valid chessboard (but some artificially rendered 3d cube (?) with a chessboard pattern), you cannot use it

please e.g. print out the one from here and take a real image with your camera

-2

943 ●9 ●6

Change this:

cv2.findChessboardCorners(gray, CHECKERBOARD, cv2.CALIB_CB_ADAPTIVE_THRESH | cv2.CALIB_CB_FAST_CHECK | cv2.CALIB_CB_NORMALIZE_IMAGE)

to:

ret, corners = cv2.findChessboardCorners(gray, CHECKERBOARD, cv2.CALIB_CB_ADAPTIVE_THRESH + cv2.CALIB_CB_FAST_CHECK + cv2.CALIB_CB_NORMALIZE_IMAGE)

Comments

supra, no, that's so silly !

berak ( 2020-08-22 01:57:59 -0600 )edit