MonoSpheres: Large-Scale Monocular SLAM-Based UAV Exploration through Perception-Coupled Mapping and Planning

2025-11-21

AI Summary

Addressing the challenges of monocular camera-based autonomous exploration in large-scale, unstructured 3D indoor and outdoor environments—namely sparse depth measurements, texture scarcity, and depth uncertainty—this paper proposes a tightly coupled mapping and planning framework. The method integrates a sparse monocular SLAM front-end with free-space oversampling, obstacle position uncertainty modeling, and disparity-driven heading control to close the perception–mapping–planning loop. It presents the first end-to-end monocular SLAM-based 3D autonomous exploration successfully deployed in real outdoor scenes. Furthermore, it introduces an uncertainty-aware fast replanning strategy, significantly improving exploration coverage and safety. Extensive evaluation across diverse real-world and simulated environments demonstrates that the system is the first open-source monocular 3D exploration solution capable of robust operation in unstructured outdoor settings.

Technology Category

• Planning, Routing, and Scheduling: Planning under Uncertainty
• Intelligent Robots: Localization, Mapping, and Navigation
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Autonomous exploration of unknown environments is a key capability for mobile robots, but it is largely unsolved for robots equipped with only a single monocular camera and no dense range sensors. In this paper, we present a novel approach to monocular vision-based exploration that can safely cover large-scale unstructured indoor and outdoor 3D environments by explicitly accounting for the properties of a sparse monocular SLAM frontend in both mapping and planning. The mapping module solves the problems of sparse depth data, free-space gaps, and large depth uncertainty by oversampling free space in texture-sparse areas and keeping track of obstacle position uncertainty. The planning module handles the added free-space uncertainty through rapid replanning and perception-aware heading control. We further show that frontier-based exploration is possible with sparse monocular depth data when parallax requirements and the possibility of textureless surfaces are taken into account. We evaluate our approach extensively in diverse real-world and simulated environments, including ablation studies. To the best of the authors' knowledge, the proposed method is the first to achieve 3D monocular exploration in real-world unstructured outdoor environments. We open-source our implementation to support future research.

Problem

• Enabling autonomous UAV exploration using only monocular cameras without dense sensors
• Addressing sparse depth data and uncertainty in monocular SLAM mapping
• Developing perception-aware planning for safe 3D exploration in unstructured environments

Innovation

• Monocular SLAM for large-scale UAV exploration
• Mapping with oversampling and uncertainty tracking
• Planning with rapid replanning and perception control