Sparse motion segmentation using multiple six-point consistencies

TL;DR

This paper introduces a fully projective motion segmentation method using six-point geometric consistency, outperforming existing affine-based approaches on standard datasets.

Contribution

The paper proposes a novel projective motion segmentation technique based on six-point geometric consistency, improving accuracy over affine models.

Findings

Outperforms state-of-the-art methods like SSC on Hopkins 155 database

Works effectively with multiple moving objects

Achieves lower maximum errors in segmentation

Abstract

We present a method for segmenting an arbitrary number of moving objects in image sequences using the geometry of 6 points in 2D to infer motion consistency. The method has been evaluated on the Hopkins 155 database and surpasses current state-of-the-art methods such as SSC, both in terms of overall performance on two and three motions but also in terms of maximum errors. The method works by finding initial clusters in the spatial domain, and then classifying each remaining point as belonging to the cluster that minimizes a motion consistency score. In contrast to most other motion segmentation methods that are based on an affine camera model, the proposed method is fully projective.