Neural Elevation Models for Terrain Mapping and Path Planning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dai, Adam, Gupta, Shubh, Gao, Grace
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916259031941120
author Dai, Adam
Gupta, Shubh
Gao, Grace
author_facet Dai, Adam
Gupta, Shubh
Gao, Grace
contents This work introduces Neural Elevations Models (NEMos), which adapt Neural Radiance Fields to a 2.5D continuous and differentiable terrain model. In contrast to traditional terrain representations such as digital elevation models, NEMos can be readily generated from imagery, a low-cost data source, and provide a lightweight representation of terrain through an implicit continuous and differentiable height field. We propose a novel method for jointly training a height field and radiance field within a NeRF framework, leveraging quantile regression. Additionally, we introduce a path planning algorithm that performs gradient-based optimization of a continuous cost function for minimizing distance, slope changes, and control effort, enabled by differentiability of the height field. We perform experiments on simulated and real-world terrain imagery, demonstrating NEMos ability to generate high-quality reconstructions and produce smoother paths compared to discrete path planning methods. Future work will explore the incorporation of features and semantics into the height field, creating a generalized terrain model.
format Preprint
id arxiv_https___arxiv_org_abs_2405_15227
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Neural Elevation Models for Terrain Mapping and Path Planning
Dai, Adam
Gupta, Shubh
Gao, Grace
Robotics
This work introduces Neural Elevations Models (NEMos), which adapt Neural Radiance Fields to a 2.5D continuous and differentiable terrain model. In contrast to traditional terrain representations such as digital elevation models, NEMos can be readily generated from imagery, a low-cost data source, and provide a lightweight representation of terrain through an implicit continuous and differentiable height field. We propose a novel method for jointly training a height field and radiance field within a NeRF framework, leveraging quantile regression. Additionally, we introduce a path planning algorithm that performs gradient-based optimization of a continuous cost function for minimizing distance, slope changes, and control effort, enabled by differentiability of the height field. We perform experiments on simulated and real-world terrain imagery, demonstrating NEMos ability to generate high-quality reconstructions and produce smoother paths compared to discrete path planning methods. Future work will explore the incorporation of features and semantics into the height field, creating a generalized terrain model.
title Neural Elevation Models for Terrain Mapping and Path Planning
topic Robotics
url https://arxiv.org/abs/2405.15227