Outlier Removal for Motion Tracking by Subspace Separation
Yasuyuki Sugaya, Kenichi Kanatani · 2002
Many feature tracking algorithms have been proposed for motion segmentation, but the resulting trajectories are not necessarily correct. In this paper, we propose a technique for removing outliers based on the knowledge that correct trajectories are constrained to be in a subspace of their domain. We first fit the subspace to the detected trajectories robustly using RANSAC and then remove those that have large residuals. Using real video sequences, we demonstrate that our method is effective even if multiple objects are moving in the scene. We also confirm that the separation accuracy is indeed improved by our method. 1