arXiv 2411.082792025-08-08cs.CVcs.RO

MBA-SLAM：运动模糊感知的高斯溅射SLAM

MBA-SLAM: Motion Blur Aware Gaussian Splatting SLAM

College of Computer Science and Technology at Zhejiang University（浙江大学计算机科学与技术学院）；School of Engineering at Westlake University（西湖大学工程学院）；School of Engineering, Westlake University（西湖大学工程学院）

Peng Wang, Lingzhe Zhao, Yin Zhang, Shiyu Zhao, Peidong Liu

备注：: Accepted to TPAMI; Deblur Gaussian Splatting SLAM

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对运动模糊下SLAM定位与建图退化问题，提出MBA-SLAM，通过运动模糊感知跟踪器与NeRF/3DGS建图器结合，建模成像过程并估计曝光轨迹，在多种数据集上超越现有方法。

摘要：

新兴的3D场景表示方法，如神经辐射场（NeRF）和3D高斯溅射（3DGS），在同时定位与建图（SLAM）中已被证明对照片级逼真渲染有效，尤其是在使用高质量视频序列作为输入时。然而，现有方法在处理运动模糊帧时存在困难，而运动模糊在现实场景中很常见，例如低光或长曝光条件下。这通常会导致相机定位精度和地图重建质量显著下降。为应对这一挑战，我们提出了一种密集视觉去模糊SLAM流程（即MBA-SLAM），以处理严重的运动模糊输入并增强图像去模糊效果。我们的方法将高效的运动模糊感知跟踪器与基于神经辐射场或高斯溅射的建图器相结合。通过精确建模运动模糊图像的物理成像过程，我们的方法同时学习3D场景表示并估计相机在曝光时间内的局部轨迹，从而主动补偿由相机运动引起的运动模糊。在我们的实验中，我们证明了MBA-SLAM在相机定位和地图重建方面均超越了以往最先进的方法，在一系列数据集（包括合成和真实数据集，涵盖清晰图像以及受运动模糊影响的图像）上展示了优越的性能，突显了我们方法的通用性和鲁棒性。代码可在https://github.com/WU-CVGL/MBA-SLAM获取。

英文摘要：

Emerging 3D scene representations, such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS), have demonstrated their effectiveness in Simultaneous Localization and Mapping (SLAM) for photo-realistic rendering, particularly when using high-quality video sequences as input. However, existing methods struggle with motion-blurred frames, which are common in real-world scenarios like low-light or long-exposure conditions. This often results in a significant reduction in both camera localization accuracy and map reconstruction quality. To address this challenge, we propose a dense visual deblur SLAM pipeline (i.e. MBA-SLAM) to handle severe motion-blurred inputs and enhance image deblurring. Our approach integrates an efficient motion blur-aware tracker with either neural radiance fields or Gaussian Splatting based mapper. By accurately modeling the physical image formation process of motion-blurred images, our method simultaneously learns 3D scene representation and estimates the cameras' local trajectory during exposure time, enabling proactive compensation for motion blur caused by camera movement. In our experiments, we demonstrate that MBA-SLAM surpasses previous state-of-the-art methods in both camera localization and map reconstruction, showcasing superior performance across a range of datasets, including synthetic and real datasets featuring sharp images as well as those affected by motion blur, highlighting the versatility and robustness of our approach. Code is available at https://github.com/WU-CVGL/MBA-SLAM.

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。