GSLAM: Initialization-robust Monocular Visual SLAM via Global Structure-from-Motion

TL;DR

GSLAM introduces a robust, fast monocular SLAM approach combining global SfM techniques with novel rank-1 matrix factorization, improving initialization robustness and loop closure accuracy.

Contribution

The paper presents a new monocular SLAM method integrating global SfM with a rank-1 matrix factorization for enhanced robustness and speed.

Findings

Achieves 4x faster reconstruction than state-of-the-art SLAM systems.

Demonstrates improved robustness to map initialization errors.

Validates effectiveness on a new ground-truth dataset and benchmark datasets.

Abstract

Many monocular visual SLAM algorithms are derived from incremental structure-from-motion (SfM) methods. This work proposes a novel monocular SLAM method which integrates recent advances made in global SfM. In particular, we present two main contributions to visual SLAM. First, we solve the visual odometry problem by a novel rank-1 matrix factorization technique which is more robust to the errors in map initialization. Second, we adopt a recent global SfM method for the pose-graph optimization, which leads to a multi-stage linear formulation and enables L1 optimization for better robustness to false loops. The combination of these two approaches generates more robust reconstruction and is significantly faster (4X) than recent state-of-the-art SLAM systems. We also present a new dataset recorded with ground truth camera motion in a Vicon motion capture room, and compare our method to prior systems on it and established benchmark datasets.