NeRFoot: Robot-Footprint Estimation for Image-Based Visual Servoing

TL;DR

This paper introduces NeRFoot, a method that uses Neural Radiance Fields to estimate robot footprints from images, enabling more precise control and safer operation in visual servoing tasks.

Contribution

It presents a novel approach combining NeRF and CNNs to accurately estimate robot footprints from visual data, improving operational safety and control.

Findings

NeRF-based footprint estimation outperforms bounding box methods.

Enhanced trajectory planning with more accurate footprint data.

Expanded safe operational area for mobile robots.

Abstract

This paper investigates the utility of Neural Radiance Fields (NeRF) models in extending the regions of operation of a mobile robot, controlled by Image-Based Visual Servoing (IBVS) via static CCTV cameras. Using NeRF as a 3D-representation prior, the robot's footprint may be extrapolated geometrically and used to train a CNN-based network to extract it online from the robot's appearance alone. The resulting footprint results in a tighter bound than a robot-wide bounding box, allowing the robot's controller to prescribe more optimal trajectories and expand its safe operational floor area.