Publication
Flow Fields: Dense Correspondence Fields for Highly Accurate Large Displacement Optical Flow Estimation
Christian Bailer; Bertram Taetz; Didier Stricker
In: IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE, 2018.
Abstract
Modern large displacement optical flow algorithms usually use an initialization by either sparse descriptor matching
techniques or dense approximate nearest neighbor fields. While the latter have the advantage of being dense, they have the major
disadvantage of being very outlier-prone as they are not designed to find the optical flow, but the visually most similar correspondence.
In this article we present a dense correspondence field approach that is much less outlier-prone and thus much better suited for optical
flow estimation than approximate nearest neighbor fields. Our approach does not require explicit regularization, smoothing (like median
filtering) or a new data term. Instead we solely rely on patch matching techniques and a novel multi-scale matching strategy. We also
present enhancements for outlier filtering. We show that our approach is better suited for large displacement optical flow estimation
than modern descriptor matching techniques. We do so by initializing EpicFlow with our approach instead of their originally used
state-of-the-art descriptor matching technique. We significantly outperform the original EpicFlow on MPI-Sintel, KITTI 2012, KITTI 2015
and Middlebury. In this extended article of our former conference publication we further improve our approach in matching accuracy as
well as runtime and present more experiments and insights.