MorphoNavi: Aerial-Ground Robot Navigation with Object Oriented Mapping in Digital Twin

2025-04-23

AI Summary

Existing aerial-ground collaborative robot navigation in unstructured environments suffers from reliance on scene-specific fine-tuning and weak coupling between multimodal perception and localization. Method: This paper proposes the first monocular-camera-driven, unified aerial-ground object-centric mapping framework. It integrates monocular visual SLAM, object-level semantic segmentation, and 3D pose estimation—enabling zero-shot generalization to diverse object categories and reconstructing their metric 3D spatial positions without per-scene adaptation. The resulting digital twin semantic map is platform-agnostic and shareable across heterogeneous robots. A dedicated aerial-ground cooperative navigation control strategy is further introduced to support autonomous localization and target search in dynamic environments. Results: Evaluated in simulated search-and-rescue missions, the MorphoGear UAV achieves real-time detection and tracking of a quadrupedal robot (<30 ms/frame), demonstrating the system’s robustness and real-time performance under complex, unstructured conditions.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

This paper presents a novel mapping approach for a universal aerial-ground robotic system utilizing a single monocular camera. The proposed system is capable of detecting a diverse range of objects and estimating their positions without requiring fine-tuning for specific environments. The system's performance was evaluated through a simulated search-and-rescue scenario, where the MorphoGear robot successfully located a robotic dog while an operator monitored the process. This work contributes to the development of intelligent, multimodal robotic systems capable of operating in unstructured environments.

Problem

• Universal aerial-ground robot navigation using monocular camera
• Object detection and position estimation without environment fine-tuning
• Intelligent multimodal robotic systems for unstructured environments

Innovation

• Monocular camera for universal aerial-ground mapping
• Object detection without environment-specific fine-tuning
• Digital twin simulation for search-and-rescue evaluation