Zhao, Zhiwen (Nanjing University of Science and Technology), Liu, Yufan (City St George's, University of London), Song, Zhenbo (Nanjing University of Science and Technology), Lu, Jianfeng (Nanjing University of Science & Technology)

Efficient 3D-Gaussian-Splatting-Based Path Planning for Ground Vehicles on Uneven Terrain

Scheduled for presentation during the Video Session "On-Demand Video Presentations" (VP-VP), Saturday, November 22, 2025, 08:00−18:00, On-Demand Platform

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on April 2, 2026

Abstract

The intricate geometric fluctuations of uneven terrains pose substantial hurdles to achieving precise terrain modeling and efficient path planning for ground vehicles. Recent breakthroughs in 3D Gaussian Splatting (3DGS) have unveiled its remarkable prowess in delivering both efficient and high-fidelity environmental representations. This study presents a novel path planning framework tailored specifically for ground vehicle navigation, which harnesses the power of 3DGS to model rugged terrains and estimate terrain attributes with remarkable conciseness and efficiency. In stark contrast to conventional point-cloud-based reconstruction techniques, our method provides a streamlined, lightweight solution without compromising structural integrity. By distilling the local geometric nuances of Gaussians, we extract critical terrain features that serve as a cornerstone for optimizing the path planning process of ground vehicles. These features empower the generation of safer, more agile trajectories across complex terrains, adeptly navigating through challenging landscapes. Comprehensive experimental validations conducted in simulated environments reveal that our approach achieves significant improvements in reconstruction efficiency, map accuracy, and planning reliability.