Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AgriGS-SLAM: Orchard Mapping Across Seasons via Multi-View Gaussian Splatting SLAM

Mirko Usuelli, David Rapado-Rincon, Gert Kootstra, Matteo Matteucci

TL;DR

AgriGS-SLAM is presented, a Visual--LiDAR SLAM framework that couples direct LiDAR odometry and loop closures with multi-camera 3D Gaussian Splatting (3DGS) rendering to deliver sharper, more stable reconstructions and steadier trajectories than recent state-of-the-art 3DGS-SLAM baselines while maintaining real-time performance on-tractor.

Abstract

Autonomous robots in orchards require real-time 3D scene understanding despite repetitive row geometry, seasonal appearance changes, and wind-driven foliage motion. We present AgriGS-SLAM, a Visual--LiDAR SLAM framework that couples direct LiDAR odometry and loop closures with multi-camera 3D Gaussian Splatting (3DGS) rendering. Batch rasterization across complementary viewpoints recovers orchard structure under occlusions, while a unified gradient-driven map lifecycle executed between keyframes preserves fine details and bounds memory. Pose refinement is guided by a probabilistic LiDAR-based depth consistency term, back-propagated through the camera projection to tighten geometry-appearance coupling. We deploy the system on a field platform in apple and pear orchards across dormancy, flowering, and harvesting, using a standardized trajectory protocol that evaluates both training-view and novel-view synthesis to reduce 3DGS overfitting in evaluation. Across seasons and sites, AgriGS-SLAM delivers sharper, more stable reconstructions and steadier trajectories than recent state-of-the-art 3DGS-SLAM baselines while maintaining real-time performance on-tractor. While demonstrated in orchard monitoring, the approach can be applied to other outdoor domains requiring robust multimodal perception.

AgriGS-SLAM: Orchard Mapping Across Seasons via Multi-View Gaussian Splatting SLAM

TL;DR

AgriGS-SLAM is presented, a Visual--LiDAR SLAM framework that couples direct LiDAR odometry and loop closures with multi-camera 3D Gaussian Splatting (3DGS) rendering to deliver sharper, more stable reconstructions and steadier trajectories than recent state-of-the-art 3DGS-SLAM baselines while maintaining real-time performance on-tractor.

Abstract

Autonomous robots in orchards require real-time 3D scene understanding despite repetitive row geometry, seasonal appearance changes, and wind-driven foliage motion. We present AgriGS-SLAM, a Visual--LiDAR SLAM framework that couples direct LiDAR odometry and loop closures with multi-camera 3D Gaussian Splatting (3DGS) rendering. Batch rasterization across complementary viewpoints recovers orchard structure under occlusions, while a unified gradient-driven map lifecycle executed between keyframes preserves fine details and bounds memory. Pose refinement is guided by a probabilistic LiDAR-based depth consistency term, back-propagated through the camera projection to tighten geometry-appearance coupling. We deploy the system on a field platform in apple and pear orchards across dormancy, flowering, and harvesting, using a standardized trajectory protocol that evaluates both training-view and novel-view synthesis to reduce 3DGS overfitting in evaluation. Across seasons and sites, AgriGS-SLAM delivers sharper, more stable reconstructions and steadier trajectories than recent state-of-the-art 3DGS-SLAM baselines while maintaining real-time performance on-tractor. While demonstrated in orchard monitoring, the approach can be applied to other outdoor domains requiring robust multimodal perception.

Paper Structure

This paper contains 21 sections, 8 equations, 3 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Proposed Approach
  4. Simultaneous Localization and Mapping
  5. SLAM Frontend
  6. SLAM Backend
  7. Multi-View 3D Gaussian Splatting
  8. Incremental Gaussian Mapping
  9. Memory Management
  10. Gaussian Splats Lifecycle Operations
  11. Optimization Loop
  12. Multimodal Loss Function
  13. Photometric Consistency
  14. Probabilistic Depth Consistency
  15. Total Multimodal Loss
  16. ...and 6 more sections

Figures (3)

  • Figure 1: The agricultural robot follows a standardized trajectory (Training- and Novel-View) in both apple and pear orchards, across dormancy, flowering, and harvesting stages. LiDAR odometry keyframes update a loop-closing factor graph that triggers background 3DGS optimization, jointly refining Gaussian submaps and camera poses. Rendered depths and images are corrected via a multimodal geometric and photometric loss, enabling simultaneous gradient-based refinement for both localization and mapping.
  • Figure 2: Agricultural platform equipped with three cameras (one horizontal and two vertical), a 32-beam LiDAR, and a VIO–GNSS–RTK system for ground truth acquisition.
  • Figure 3: Comparison of SLAM trajectories against ground truth.