Collection of different motion estimation algorithms
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
data
tools
.gitmodules
README.txt
example.m
motionEstimatorClass.m
testMotionEstimatorClass.m
testMotionEstimatorClass3D.m

README.txt

% MotionEstimation GUI 
%
% Author Information: 
% Hendrik Dirks
% Institute for Computational and Applied Mathematics
% University of Muenster, Germany
%
% Contact: hendrik.dirks@wwu.de
%
%
% Version 2.0
% Date: 2016-09-27

% All Rights Reserved
%
% Permission to use, copy, modify, and distribute this software and its
% documentation for any purpose other than its incorporation into a
% commercial product is hereby granted without fee, provided that the
% above copyright notice appear in all copies and that both that
% copyright notice and this permission notice appear in supporting
% documentation, and that the name of the author and University of Muenster not be used in
% advertising or publicity pertaining to distribution of the software
% without specific, written prior permission.

If you use this framwork, please use the following citation
@PhdThesis{Dir15,
  author       = {Dirks, H.},
  title        = {Variational Methods for Joint Motion Estimation and Image Reconstruction},
  school       = {Institute for Computational and Applied Mathematics University of Muenster},
  month        = {june},
  year         = {2015},
  keywords     = {Image Processing, Image Reconstruction, Motion Estimation, Joint Image Reconstruction and Motion Estimation, Total Variation, Optical Flow, Microscopy Imaging, Biomedical Imaging, Temporal Inpainting, Variational Methods, Primal-Dual Methods},
  url          = \{/2015/Dir15},
}

Install Instructions:
Make sure you have downloaded a copy of flexBox from www.flexbox.im and this copy is in your MATALB path. 
Moreover, you should compile the C++-module for runtime reasons! Afterwards, run example.m

Further copyright notes:
- The files "computeColor.m", "writeFlowFile.m" and the original version of "flowToColorV2.m" have been developed by Deqing Sun, Department of Computer Science, Brown University (dqsun@cs.brown.edu)
- The dataset in the "data" folder comes from the Middlebury optical flow database (http://vision.middlebury.edu/flow/)

List of changes:
## Version 2.0 ##
- Motion estimator completely rewritten:
-- Based on flexBox
-- CUDA and C++ support
-- Large-scale motion estimation based on coarse-to-fine warping
## Version 1.4 ##
- Bugfixes
## Version 1.3 ##
- Bugfixes
- Missing files added
## Version 1.2 ##
- Bugfixes
## Version 1.1 ##
- Add a coarse-to-fine warping for all algorithms and set the maxIterations to default number of 10. Coarse-to-fine warping can be turned off by setting the "Pyramid steps" 
  to 1 and increasing the "Max Iterations"
- Added a L1-TV optical flow method with nonlinear data attachment term (in fact it's still a linearization, but closer to the classical attachment term)