Motion Detail Preserving Optical Flow Estimation

Li Xu     Jiaya Jia      Yasuyuki Matsushita

Abstract

A common problem of optical flow estimation in the multi-scale variational framework is that fine motion structures cannot always be correctly estimated, especially for regions with significant and abrupt displacement variation. A novel extended coarse-to-fine (EC2F) refinement framework is introduced in this paper to address this issue, which reduces the reliance of flow estimates on their initial values propagated from the coarse level and enables recovering many motion details in each scale. The contribution of this paper also includes adaptation of the objective function to handle outliers and development of a new optimization procedure. The effectiveness of our algorithm is demonstrated using the Middlebury optical flow benchmark and by experiments on challenging examples that involve large-displacement motion.

Downloads

[CVPR 2010 Paper], [TPAMI 2012 Paper], [Talk Slides]

Optical Flow Estimation Software


Our Related Work

SegOF: A Segmentation Based Variational Model for Accurate Optical Flow Estimation (ECCV 2008) (Software)

SIOF: Scale Invariant Optical Flow (ECCV 2012)

Consistent Binocular Depth and Scene Flow with Chained Temporal Profiles (IJCV 2012)


More Examples

HumanEvaII #0 HumanEvaII #1 lambda=8
Baseball #0 Baseball #1 lambda=8
Girl #0 Girl #1 lambda=12
Football #0 Football #1 lambda=20