English

A Continuous Optimization Approach for Efficient and Accurate Scene Flow

Computer Vision and Pattern Recognition 2016-07-28 v1

Abstract

We propose a continuous optimization method for solving dense 3D scene flow problems from stereo imagery. As in recent work, we represent the dynamic 3D scene as a collection of rigidly moving planar segments. The scene flow problem then becomes the joint estimation of pixel-to-segment assignment, 3D position, normal vector and rigid motion parameters for each segment, leading to a complex and expensive discrete-continuous optimization problem. In contrast, we propose a purely continuous formulation which can be solved more efficiently. Using a fine superpixel segmentation that is fixed a-priori, we propose a factor graph formulation that decomposes the problem into photometric, geometric, and smoothing constraints. We initialize the solution with a novel, high-quality initialization method, then independently refine the geometry and motion of the scene, and finally perform a global non-linear refinement using Levenberg-Marquardt. We evaluate our method in the challenging KITTI Scene Flow benchmark, ranking in third position, while being 3 to 30 times faster than the top competitors.

Keywords

Cite

@article{arxiv.1607.07983,
  title  = {A Continuous Optimization Approach for Efficient and Accurate Scene Flow},
  author = {Zhaoyang Lv and Chris Beall and Pablo F. Alcantarilla and Fuxin Li and Zsolt Kira and Frank Dellaert},
  journal= {arXiv preprint arXiv:1607.07983},
  year   = {2016}
}

Comments

Accepted in ECCV 2016. Please refer to the ECCV16-springer for detailed information

R2 v1 2026-06-22T15:05:21.187Z