Trajectory Comparison: ORB-SLAM3 vs. pySLAM

This repository contains code and data for comparing trajectories generated by two different SLAM systems: ORB-SLAM3 and pySLAM. Specifically, it focuses on computing the Absolute Pose Error (APE) and Relative Pose Error (RPE) of trajectories collected from two environments.

Files

• MH01_GT.tum: EuRoC ground truth trajectory in TUM format.
• f_ORBSLAM3_EuRoC_TUM_Trajectory.tum: Trajectory generated by ORB-SLAM3 for the EuRoC dataset.
• f_pySLAM_EuRoC_TUM_Trajectory.tum: Trajectory generated by pySLAM for the EuRoC dataset.
• trajectories_comparison.ipynb: Jupyter Notebook containing code to plot trajectories with ground truth.

Trajectory Collection

The pySLAM trajectories were collected using my implementation of trajectory saving, available here. This implementation supports trajectory saving in three different formats: KITTI, TUM, and EuRoC.

Dependencies

• evo: A Python package for pose evaluation. Install it via pip:

pip install evo

Usage

1. Clone the repository:

git clone https://github.com/anathonic/Trajectory-Comparison-ORB-SLAM3-pySLAM.git