Inhalt des Dokuments
Robust Local Optical Flow
The presented work is motivated by the problem of local motion estimation via robust regression with linear models. In order to increase the robustness of the motion estimates we propose a novel Robust Local Optical Flow approach based on a modified Hampel estimator. We show the deficiencies of the least squares estimator used by the standard KLT tracker when the assumptions made by Lucas/Kanade are violated. We propose a strategy to adapt the window sizes to cope with the Generalized Aperture Problem...
- Tobias Senst, Volker Eiselein, Thomas Sikora
Robust Local Optical Flow for Feature Tracking
IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), IEEE, vol. 22, no. 9, September 2012, pp. 1377--1387
- Tobias Senst, Volker Eiselein, Rubén Heras Evangelio, Thomas Sikora
Robust Modified L2 Local Optical Flow Estimation and Feature Tracking
IEEE Workshop on Motion and Video Computing (WMVC), Kona, USA, 05.01.2011 - 07.01.2011, pp. 685--690
IEEE Catalog Number: CFP11082-CDR ISBN: 978-1-4244-9495-8
|stay connected via Twitter||twitter.com/RLOFLib|
08/01/2013 Release Version 1.1
- Improved GPU implementation by extending image borders and enhanced the runtime performance of the CL kernels.
- Parallelize CPU implementation using Thread Building Blocks and SSE2.
We provide binaries of the RLOF feature tracker in order to help other researchers to compare their results or to use our work as a module for their research. The files contain a binary package for the Windows operating system.
The documentation is included in the RLOF package (/Doc/html/index.html) or available online by the following link:
Current versions are available:
- Windows Visual Studio 2008 [x86/x64]
- Windows Visual Studio 2010 [x86/x64]
- Windows Matlab [x86/x64]
To receive the code, please fill out the form.
All code is provided for research purposes only and without any warranty. Any commercial use is prohibited. By using the code in your research work, you should cite the respective paper.