一种从粗略姿态图像和带噪激光雷达点云进行高斯溅射的约束优化方法
A Constrained Optimization Approach for Gaussian Splatting from Coarsely-posed Images and Noisy Lidar Point Clouds
AI总结：
针对3D高斯溅射需SfM初始化的问题，提出约束优化方法，分解相机姿态为两步优化并引入参数与几何约束，无需SfM即可同时估计相机姿态与重建3D场景，性能优于现有基线及COLMAP辅助方法。
摘要：
3D高斯溅射（3DGS）是一种强大的重建技术，但需要从精确的相机姿态和高保真点云初始化。通常初始化来自运动恢复结构（SfM）算法；然而SfM耗时，限制了3DGS在现实场景和大规模场景重建中的应用。我们提出一种无需SfM支持的约束优化方法，用于同时进行相机姿态估计和3D重建。方法核心是将相机姿态分解为相机到（设备）中心、（设备）中心到世界的序列优化。为此，我们提出两个基于参数组敏感性的优化约束，限制各参数搜索空间。此外，由于直接从带噪点云学习场景几何，我们提出几何约束以提升重建质量。实验表明，在我们收集的数据集和两个公开基准上，该方法显著优于现有（多模态）3DGS基线及COLMAP辅助方法。
英文摘要：
3D Gaussian Splatting (3DGS) is a powerful reconstruction technique, but it needs to be initialized from accurate camera poses and high-fidelity point clouds. Typically, the initialization is taken from Structure-from-Motion (SfM) algorithms; however, SfM is time-consuming and restricts the application of 3DGS in real-world scenarios and large-scale scene reconstruction. We introduce a constrained optimization method for simultaneous camera pose estimation and 3D reconstruction that does not require SfM support. Core to our approach is decomposing a camera pose into a sequence of camera-to-(device-)center and (device-)center-to-world optimizations. To facilitate, we propose two optimization constraints conditioned to the sensitivity of each parameter group and restricts each parameter's search space. In addition, as we learn the scene geometry directly from the noisy point clouds, we propose geometric constraints to improve the reconstruction quality. Experiments demonstrate that the proposed method significantly outperforms the existing (multi-modal) 3DGS baseline and methods supplemented by COLMAP on both our collected dataset and two public benchmarks.
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。
