DynORecon: Dynamic Object Reconstruction for Navigation

TL;DR

DynORecon is a real-time system that reconstructs moving objects in 3D to aid navigation, improving accuracy and efficiency by leveraging Dynamic SLAM for dynamic scene understanding.

Contribution

It introduces a novel dynamic object reconstruction method that integrates with Dynamic SLAM, enabling accurate, real-time 3D mapping of moving objects for navigation.

Findings

Achieves ~10 cm accuracy in object reconstruction.

Operates at approximately 20 frames per second.

Effective on both simulated and real-world outdoor datasets.

Abstract

This paper presents DynORecon, a Dynamic Object Reconstruction system that leverages the information provided by Dynamic SLAM to simultaneously generate a volumetric map of observed moving entities while estimating free space to support navigation. By capitalising on the motion estimations provided by Dynamic SLAM, DynORecon continuously refines the representation of dynamic objects to eliminate residual artefacts from past observations and incrementally reconstructs each object, seamlessly integrating new observations to capture previously unseen structures. Our system is highly efficient (~20 FPS) and produces accurate (~10 cm) reconstructions of dynamic objects using simulated and real-world outdoor datasets.