A New Approach to Two-View Motion Segmentation Using Global Dimension Minimization

摘要

We present a new approach to rigid-body motion segmentation from two views.We use a previously developed nonlinear embedding of two-view pointcorrespondences into a 9-dimensional space and identify the different motionsby segmenting lower-dimensional subspaces. In order to overcome nonuniformdistributions along the subspaces, whose dimensions are unknown, we suggest thenovel concept of global dimension and its minimization for clustering subspaceswith some theoretical motivation. We propose a fast projected gradientalgorithm for minimizing global dimension and thus segmenting motions from2-views. We develop an outlier detection framework around the proposed method,and we present state-of-the-art results on outlier-free and outlier-corruptedtwo-view data for segmenting motion.