ORB-SLAM: a Versatile and Accurate Monocular SLAM System

TL;DR

ORB-SLAM is a real-time, versatile monocular SLAM system that achieves high robustness and accuracy across diverse environments by using a unified feature-based approach for all SLAM tasks.

Contribution

The paper introduces ORB-SLAM, a novel monocular SLAM system that integrates tracking, mapping, relocalization, and loop closing using the same features, with automatic initialization and lifelong operation capabilities.

Findings

Achieves state-of-the-art accuracy on multiple datasets.

Operates in real time in various environments.

Provides publicly available source code.

Abstract

This paper presents ORB-SLAM, a feature-based monocular SLAM system that operates in real time, in small and large, indoor and outdoor environments. The system is robust to severe motion clutter, allows wide baseline loop closing and relocalization, and includes full automatic initialization. Building on excellent algorithms of recent years, we designed from scratch a novel system that uses the same features for all SLAM tasks: tracking, mapping, relocalization, and loop closing. A survival of the fittest strategy that selects the points and keyframes of the reconstruction leads to excellent robustness and generates a compact and trackable map that only grows if the scene content changes, allowing lifelong operation. We present an exhaustive evaluation in 27 sequences from the most popular datasets. ORB-SLAM achieves unprecedented performance with respect to other state-of-the-art monocular SLAM approaches. For the benefit of the community, we make the source code public.