Original Loop-closure Detection Algorithm for Monocular vSLAM

TL;DR

This paper introduces a novel loop-closure detection algorithm for monocular vSLAM that enhances mapping accuracy and enables real-time drone navigation.

Contribution

The paper presents an original loop-closure detection algorithm adaptable to various vSLAM methods, improving accuracy and real-time performance.

Findings

Improved mapping accuracy in monocular vSLAM.

Reduced computational time enabling real-time drone control.

Compatible with dense, semi-dense, and feature-based methods.

Abstract

Vision-based simultaneous localization and mapping (vSLAM) is a well-established problem in mobile robotics and monocular vSLAM is one of the most challenging variations of that problem nowadays. In this work we study one of the core post-processing optimization mechanisms in vSLAM, e.g. loop-closure detection. We analyze the existing methods and propose original algorithm for loop-closure detection, which is suitable for dense, semi-dense and feature-based vSLAM methods. We evaluate the algorithm experimentally and show that it contribute to more accurate mapping while speeding up the monocular vSLAM pipeline to the extent the latter can be used in real-time for controlling small multi-rotor vehicle (drone).