###################################################################
###################################################################
ABOUT: This software implements our approach to optical flow estimation [1] with several data cost functions.
The additional and optional library
- Eigen is not included.
To download that package follow the link: http://eigen.tuxfamily.org/index.php?title=Main_Page and read the licensing information provided there.
========================================================================== DISCLAIMER: This demo software has been rewritten for the sake of simplifying the implementation. Therefore, the results produced by the code may differ from those presented in the papers [1]. In fact the results should be better on the KITTI dataset: http://www.cvlibs.net/datasets/kitti/.
==========================================================================
IMPORTANT: If you use this software you should cite the following in any resulting publication:
[1] An Evaluation of Data Costs for Optical Flow
C. Vogel, S. Roth and K. Schindler
In GCPR, Saarbruecken, Germany, September 2013
INSTALLING & RUNNING
-
(Optional) Download and install eigen from http://eigen.tuxfamily.org/index.php?title=Main_Page and place it into the folder ./Source. Alternatively one can change the switch in the file compileMex to compile a slightly slower standalone version.
-
Start MATLAB and run compileMex.m to build the utilities binaries. (This step can be omitted if you are using Windows 64 bit or Unix 64 bit And do not want ot use OpenMP.) Adjust the compiler flags accordingly for your purposes (defaults should work in most cases).
-
From folder DataFlow run calltest( xx ) - example given as comment in the code. This will execute a KITTI example.
-
Otherwise load images I1, I2 and run : flow = Data_flow(1.25/255, 12.333, 3, 0.9, I1, I2, 10, 2, 1, 0, 0.5, 16, 0, 1) 'flow' contains the computed 2d flow as usual. Here I1, I2 are the input images (gray-level only so far). All parameter are explained in the script, Data_flow.
Note that the code should perform slightly better as published.
CHANGES 1.0 April 19, 2014 Initial public release