Robust Optical Flow Estimation
Javier Sánchez Pérez, Nelson Monzón López, Agustín Salgado de la Nuez
Javier Sánchez Pérez, Nelson Monzón López, and Agustín Salgado de la Nuez, Robust Optical Flow Estimation, Image Processing On Line, 3 (2013), pp. 252–270.

Communicated by Enric Meinhardt-Llopis
Demo edited by Enric Meinhardt-Llopis


In this work, we describe an implementation of the variational method proposed by Brox etal. in 2004, which yields accurate optical flows with low running times. It has several benefitswith respect to the method of Horn and Schunck: it is more robust to the presence of outliers,produces piecewise-smooth flow fields and can cope with constant brightness changes. Thismethod relies on the brightness and gradient constancy assumptions, using the information ofthe image intensities and the image gradients to find correspondences. It also generalizes theuse of continuous L1 functionals, which help mitigate the effect of outliers and create a TotalVariation (TV) regularization. Additionally, it introduces a simple temporal regularizationscheme that enforces a continuous temporal coherence of the flow fields.


Supplementary Materials

Non-Reviewed Supplementary Materials

These files and information are provided by the authors and have not been reviewed.