MDP-based 3-D Motion Segmentation

Abstract

This paper presents a novel motion segmentation algorithm on the basis of mixture of Dirichlet process (MDP) models, a kind of nonparametric Bayesian framework. In contrast to previous approaches, our algorithm considers two-view motion segmentation and its model selection regarding to the number of motion models as an indivisible problem. The proposed algorithm can simultaneously infer the number of motion models, estimate the cluster memberships of correspondence points, and identify the outliers of input data. The key idea is to use MDP models to fully exploit the epipolar constraints before making premature decisions about the number of motion models. By working on a discretized motion parameter space, our algorithm does not suffer from the sampling problems encountered in RANSAC-based approaches. In the experiments, we compare the proposed algorithm with subspace separation, naive RANSAC and GPCA on both synthetic data and real image data. The experimental results show that we can handle more motions and have satisfactory performance in the presence of various levels of noise and outlier.

Reference

Yong-Dian Jian and Chu-Song Chen
Two-View Motion Segmentation by Mixtures of Dirichlet Process with Model Selection and Outlier Removal [Paper PDF] [Poster PDF]
Proceedings of 11th International Conference on Computer Vision, Rio de Janeiro, Brazil, 2007 (ICCV2007)

Yong-Dian Jian and Chu-Song Chen
Motion Segmentation with Model Selection and Outlier Removal by RANSAC-Enhanced Dirichlet Process Mixture Models
Submitted to IEEE Transactions on Pattern Analysis and Machine Intelligence, under first round review process

Poster