amcctoolbox / Home
This toolbox is the Automatic Multi-Camera Calibration (AMCC) Toolbox, written for Matlab. From a set of checkerboard images from multiple fixed cameras with overlapping views, the toolbox can calibrate both intrinsics and extrinsics with no user input. No need for manual clicking or manual optimization! It takes advantage of parallel processing on multiple cores to speed up computation where possible.
The toolbox is a modified version of Bouget's Camera Calibration Toolbox for Matlab and the Robust Automatic Detection Of Calibration Chessboards (RADOCC) toolbox. The toolbox extends automatic checkerboard detection to multiple cameras and implements entirely automated monocular and stereo calibration procedures.
Get The Toolbox
Make sure you have the Matlab Image Processing Toolbox installed before continuing.
To get the latest updates (and some potential bugs) checkout the toolbox anonymously via git:
$ git clone https://bitbucket.org/michaeldwarren/amcctoolbox.git
Using The Toolbox
Once you have downloaded the software, check out the User Guide.
Referencing The Toolbox
If you use the toolbox in the production of a paper, it would be appreciated if you could cite the following paper in your own work:
M. Warren, D. McKinnon, B. Upcroft, "Online Calibration of Stereo Rigs for Long-Term Autonomy", in International Conference on Robotics and Automation, Karlsruhe, Germany, 2013. (PDF) (BibTex)
Don't have a dataset to experiment with? Download this example checkerboard dataset with four cameras to calibrate: Example_Multicamera_Calibration_Dataset.zip (280.8MB)
Need help? Bugs, crashes or other issues?
michaeldwarren AT gmail DOT com. Nine times out of ten your setup script will have a misspelled string or other minor setup error. Double check your script before looking for further help. If you really need it, please submit your verbatim script and a few example images from your calibration dataset. Specify the toolbox version number and describe in as much detail as possible the issue you are having.