4
$\begingroup$

I am looking for a library for disparity map / stereo vision computation. These are my requirements:

  1. C++
  2. Multi-platform (Linux, Windows, OSX)
  3. (preferrable but not mandatory) not CUDA based
  4. Suited for robotics (e.g. it should work even if the images are not perfectly rectified and the cameras are not perfectly calibrated)
  5. Suitable for tracking purposes (20fps or more)
  6. Performing even with low-res images (e.g. 320x240px)
  7. Open Source
$\endgroup$

closed as primarily opinion-based by TobiasK, Andrew May 20 '15 at 6:09

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • $\begingroup$ Have you looked at OpenCV? $\endgroup$ – CroCo May 13 '15 at 17:48
  • $\begingroup$ Yes. OpenCV was our first choice but it falls short mainly because of item #4. Further, it is very slow (running at 1fps with images at 640x480px even on powerful machines). $\endgroup$ – alecive May 14 '15 at 11:40
3
$\begingroup$

I'll try to answer this question on my own, but feel free to add more information on that.

We tried OpenCV's standard disparity function, but it fell short mainly because of item #4. Further, it is very slow (running at 1fps with images at 640x480px even on powerful machines).

Recently, we are doing some tests with libelas and it seems promising. It is more robust, faster and it generally works better.

$\endgroup$
0
$\begingroup$

I just learned about libviso. I didn't compare to all of your requirements, but it looks pretty good.

$\endgroup$
  • 2
    $\begingroup$ Both libelas and libviso libraries are from the same author; the former deals with the disparity, whereas the latter with the visual odometry. $\endgroup$ – Ugo May 14 '15 at 15:06

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