VIO-Aided Structure from Motion Under Challenging Environments

TL;DR

This paper introduces a robust Structure from Motion pipeline that integrates visual-inertial odometry to improve 3D reconstruction accuracy and robustness in challenging environments, outperforming existing methods.

Contribution

It proposes a geometric verification method and a scalable reconstruction approach that leverage local odometry for enhanced SfM performance.

Findings

Outperforms state-of-the-art SfM methods in accuracy

Demonstrates robustness in challenging environments

Efficient and scalable reconstruction process

Abstract

In this paper, we present a robust and efficient Structure from Motion pipeline for accurate 3D reconstruction under challenging environments by leveraging the camera pose information from a visual-inertial odometry. Specifically, we propose a geometric verification method to filter out mismatches by considering the prior geometric configuration of candidate image pairs. Furthermore, we introduce an efficient and scalable reconstruction approach that relies on batched image registration and robust bundle adjustment, both leveraging the reliable local odometry estimation. Extensive experimental results show that our pipeline performs better than the state-of-the-art SfM approaches in terms of reconstruction accuracy and robustness for challenging sequential image collections.