Abstract:Structure from motion (SfM) is a prevalent solution for vision measurement. Traditional SfM methods rely on sparse feature matching to establish visual associations between different views. However, both the incremental-based SfM and the global-based SfM show sensitiveness to outliers generated by existing handcrafted features under challenging environments. Planar marker-assisted SfM is an alternative solution that employs artificial landmarks to help robust pose estimation. Nevertheless, using planar markers also exists the planar pose ambiguity problem, which means there may be more than one reasonable solution for pose estimation; it is a great challenge for camera pose estimation. This article proposes an efficient planar marker-assisted global SfM method and resolves the planar pose ambiguity problem by examining the rotation consistency across multiple views. In particular, we construct two complementary graphs and present a graph filter method to filter out the views with pose ambiguity according to our designed rotation consistency checking. Afterward, we perform sparse feature matching following the proposed graph structure and make a pose-fixed bundle adjustment for high-quality 3-D reconstruction. Extensive experiments have been conducted on synthetic and real datasets, and the results verify the effectiveness of the proposed method. Our method outperforms several state-of-the-art methods with encouraging efficiency, especially in challenging scenes.

Structure from Motion: Sparse Versus Dense Correspondence Methods

Efficient Non-Consecutive Feature Tracking for Robust Structure-From-Motion

TC-SfM: Robust Track-Community-Based Structure-from-Motion

Scalable Dense Non-rigid Structure-from-Motion: A Grassmannian Perspective

EC-SfM: Efficient Covisibility-based Structure-from-Motion for Both Sequential and Unordered Images

Nonrigid Structure from Motion Via Sparse Representation.

Dense Non-Rigid Structure from Motion: A Manifold Viewpoint

Efficient and Robust Large-Scale Structure-from-motion Via Track Selection and Camera Prioritization

Efficient Non-Consecutive Feature Tracking For Structure-From-Motion

Structure from Motion Using Sequential Monte Carlo Methods

Dense Non-Rigid Structure-from-motion Made Easy — A Spatial-Temporal Smoothness Based Solution

CSfM: Community-based Structure from Motion

Efficient Covisibility-based Image Matching for Large-Scale SfM

Non-rigid Dense Correspondence with Applications for Image Enhancement

Parallel Structure from Motion for Sparse Point Cloud Generation in Large-Scale Scenes

Global Structure-from-Motion Revisited

Global Structure-from-Motion by Similarity Averaging

Multi-Body Non-Rigid Structure-from-Motion.

An Efficient Visual SfM Framework Using Planar Markers.

Geometry-aware Feature Matching for Large-Scale Structure from Motion

MASt3R-SfM: a Fully-Integrated Solution for Unconstrained Structure-from-Motion