ORB-SLAM3AB: Augmenting ORB-SLAM3 to Counteract Bumps with Optical Flow Inter-frame Matching

TL;DR

An enhancement to the ORB-SLAM3 algorithm, tailored for applications on rugged road surfaces, which adeptly combines feature point matching with optical flow methods, capitalizing on the high robustness of optical flow in complex terrains and the high precision of feature points on smooth surfaces.

Abstract

This paper proposes an enhancement to the ORB-SLAM3 algorithm, tailored for applications on rugged road surfaces. Our improved algorithm adeptly combines feature point matching with optical flow methods, capitalizing on the high robustness of optical flow in complex terrains and the high precision of feature points on smooth surfaces. By refining the inter-frame matching logic of ORB-SLAM3, we have addressed the issue of frame matching loss on uneven roads. To prevent a decrease in accuracy, an adaptive matching mechanism has been incorporated, which increases the reliance on optical flow points during periods of high vibration, thereby effectively maintaining SLAM precision. Furthermore, due to the scarcity of multi-sensor datasets suitable for environments with bumpy roads or speed bumps, we have collected LiDAR and camera data from such settings. Our enhanced algorithm, ORB-SLAM3AB, was then benchmarked against several advanced open-source SLAM algorithms that rely solely on laser or visual data. Through the analysis of Absolute Trajectory Error (ATE) and Relative Pose Error (RPE) metrics, our results demonstrate that ORB-SLAM3AB achieves superior robustness and accuracy on rugged road surfaces.

Figures (3)

• Figure 1: We improve visual SLAM robustness by converting frames to grayscale, then processing them with ORB extraction and optical flow. Gaussian filtering reduces noise before applying optical flow. To maintain ORB feature accuracy, we start with a minimal point selection for optical flow, increasing it only with significant camera shake. The green box represents the original ORB-SLAM3 framework.
• Figure 2: The figure represents our self-collected dataset of speed bump and bumpy road sections, with the speed bumps and bumpy areas highlighted in red boxes.
• Figure 4: We tested the collected data under various weather, lighting, and road conditions using three laser SLAM algorithms and two visual SLAM algorithms. In high-speed snow conditions with speed bumps, the ORB-SLAM3 algorithm exhibited severe trajectory errors due to intense vibrations. In high-speed snowy night conditions with speed bumps, the DSO-SLAM trajectory was also far from ideal due to the lack of loop closure detection, among other reasons. Our proposed ORB-SLAM3AB algorithm notably outperformed the original ORB-SLAM3 and DSO-SLAM in terms of map accuracy on speed bump surfaces.