arXiv 2409.130552025-11-08cs.ROcs.CV

MGSO：基于高效3D高斯泼溅的单目实时光度SLAM

MGSO: Monocular Real-time Photometric SLAM with Efficient 3D Gaussian Splatting

Vision and Image Processing Lab at the Faculty of System Design Engineering, at the University of Waterloo（视觉与图像处理实验室，系统设计工程学院，滑铁卢大学）；Vision and Robotics Lab, Maroun Semaan Faculty of Engineering and Architecture, American University of Beirut（视觉与机器人实验室，马鲁恩·塞马安工程与建筑学院，贝鲁特美国大学）

Yan Song Hu, Nicolas Abboud, Muhammad Qasim Ali, Adam Srebrnjak Yang, Imad Elhajj, Daniel Asmar, Yuhao Chen, John S. Zelek

备注：: This is the pre-print version of a work that has been published in ICRA 2025 with doi: 10.1109/ICRA55743.2025.11127380. This version may no longer be accessible without notice. Copyright 2025 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses. Please cite the official version
发表信息：: 2025 IEEE International Conference on Robotics and Automation (ICRA), Atlanta, GA, USA, 2025, pp. 11061-11067
DOI：: 10.1109/ICRA55743.2025.11127380

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出MGSO系统，将光度SLAM与3D高斯泼溅结合，利用SLAM提供的稠密点云初始化高斯，在RGB输入下实现质量、内存和速度均衡的实时单目稠密重建，并在多个数据集上超越现有方法。

摘要：

具有稠密三维建图的实时SLAM在计算上具有挑战性，尤其是在资源受限的设备上。近年来3D高斯泼溅（3DGS）的发展为实时稠密三维重建提供了一种有前景的方法。然而，现有的基于3DGS的SLAM系统难以在硬件简洁性、速度和地图质量之间取得平衡。大多数系统在上述方面中仅在一两个方面表现出色，但很少能同时实现全部目标。一个关键问题是在同时进行SLAM的过程中初始化3D高斯的困难。为了解决这些挑战，我们提出了Monocular GSO（MGSO），这是一种将光度SLAM与3DGS相结合的新型实时SLAM系统。光度SLAM为3DGS初始化提供稠密的结构化点云，从而加速优化，并以更少的高斯生成更高效的地图。实验结果表明，我们的系统生成的重建结果在质量、内存效率和速度之间取得了平衡，优于当前最先进的方法。此外，我们的系统仅使用RGB输入即可实现所有结果。我们在Replica、TUM-RGBD和EuRoC数据集上对当前的实时稠密重建系统进行了评估。我们不仅超越了现有系统，而且实验还表明我们在笔记本电脑硬件上保持了性能，使其成为机器人、A/R以及其他实时应用的实用解决方案。

英文摘要：

Real-time SLAM with dense 3D mapping is computationally challenging, especially on resource-limited devices. The recent development of 3D Gaussian Splatting (3DGS) offers a promising approach for real-time dense 3D reconstruction. However, existing 3DGS-based SLAM systems struggle to balance hardware simplicity, speed, and map quality. Most systems excel in one or two of the aforementioned aspects but rarely achieve all. A key issue is the difficulty of initializing 3D Gaussians while concurrently conducting SLAM. To address these challenges, we present Monocular GSO (MGSO), a novel real-time SLAM system that integrates photometric SLAM with 3DGS. Photometric SLAM provides dense structured point clouds for 3DGS initialization, accelerating optimization and producing more efficient maps with fewer Gaussians. As a result, experiments show that our system generates reconstructions with a balance of quality, memory efficiency, and speed that outperforms the state-of-the-art. Furthermore, our system achieves all results using RGB inputs. We evaluate the Replica, TUM-RGBD, and EuRoC datasets against current live dense reconstruction systems. Not only do we surpass contemporary systems, but experiments also show that we maintain our performance on laptop hardware, making it a practical solution for robotics, A/R, and other real-time applications.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。