Enhancing exploration algorithms for navigation with visual SLAM

TL;DR

This paper improves exploration algorithms for autonomous navigation by integrating them with visual SLAM, and evaluates their performance using realistic simulation with different depth inputs.

Contribution

It introduces enhancements for exploration algorithms tailored for vision-based SLAM systems and assesses their effectiveness in simulated environments.

Findings

Enhanced algorithms improve exploration coverage.

Evaluation shows robustness with different depth inputs.

Simulation results validate the proposed methods.

Abstract

Exploration is an important step in autonomous navigation of robotic systems. In this paper we introduce a series of enhancements for exploration algorithms in order to use them with vision-based simultaneous localization and mapping (vSLAM) methods. We evaluate developed approaches in photo-realistic simulator in two modes: with ground-truth depths and neural network reconstructed depth maps as vSLAM input. We evaluate standard metrics in order to estimate exploration coverage.