0
$\begingroup$

For stereo cameras on the market, two cameras are always mounted side by side and with a displacement that is perpendicular to the cameras’ optical axes. I take this setup for granted. One idea came to my mind whether this is necessary? If two cameras are not parallel and have different focal length, camera calibration can correct the difference. Why are two cameras mounted in parallel? My guess is that two cameras can have a large overlapping region. Am I correct?

$\endgroup$
  • 1
    $\begingroup$ Because that's the configuration you'd go for if your recording a stereo image for playback on a left and right channel for the human eyes, which happen to have the same configuration, maybe? $\endgroup$ – Marcus Müller Feb 23 '17 at 7:21
  • $\begingroup$ I understand. I don't use the images for playback but rather for depth calculation. How can I evaluate the calibration performance If I calibrate one stereo camera? Will it not be a problem if two cameras are not mounted in parallel exactly? $\endgroup$ – Jogging Song Feb 23 '17 at 7:36
  • $\begingroup$ If they were mounted in series, the camera that is behind would only see the backside of the camera in front. $\endgroup$ – Dilip Sarwate Mar 25 '17 at 16:10
1
$\begingroup$

When you say cameras are parallel it only means that the center pixels are parallel.

Every pixel has a different angle, and that refers to the overlapping regions you were talking about. Actually there is at most one overlapping voxel (3d pixel kind of) for every 2 pixels!

There is more about it in epipolar geomtry , it's interesting and pretty simple!

$\endgroup$
  • $\begingroup$ I mean that the optical axes of the two cameras should be parallel. The overlap of two cameras after rectification is large. $\endgroup$ – Jogging Song Feb 23 '17 at 8:29
  • 1
    $\begingroup$ If your question is why the cameras are specifically parallel then there are 2 answers. First it's because this way it is very similar to the human case (eyes being parallel approximately), and secondly there is a lot of overlapping and you can see very deep. Note that you can only get depth of the overlapping area, and therefor to get depth of far objects the cameras should be approximately parallel. $\endgroup$ – Cherny Feb 23 '17 at 8:33
  • $\begingroup$ Your comment is very helpful. How can I evaluate the calibration performance If I calibrate one stereo camera? Will it not be a problem if two cameras are not mounted in parallel exactly? $\endgroup$ – Jogging Song Feb 23 '17 at 8:54
  • $\begingroup$ I am sorry but I know these field only through theoretical eyes. I am afraid to give you answers that won't be correct enough. If you have something more specific maybe I can help $\endgroup$ – Cherny Feb 23 '17 at 9:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.