Robust Large-Scale Localization in 3D Point Clouds Revisited

TL;DR

This paper re-evaluates existing algorithms for 6-DOF pose estimation within 3D point clouds, highlighting their strengths and weaknesses, and introduces improvements in point selection, RANSAC, and pose quality estimation.

Contribution

It provides a critical analysis of previous methods and proposes new priors and parameters to enhance pose estimation accuracy and robustness.

Findings

Algorithms can compute poses from 3 or 4 points with known or unknown focal length.

Identified advantages and shortcomings of existing methods.

Proposed additional priors and parameters improve pose estimation.

Abstract

We tackle the problem of getting a full 6-DOF pose estimation of a query image inside a given point cloud. This technical report re-evaluates the algorithms proposed by Y. Li et al. "Worldwide Pose Estimation using 3D Point Cloud". Our code computes poses from 3 or 4 points, with both known and unknown focal length. The results can easily be displayed and analyzed with Meshlab. We found both advantages and shortcomings of the methods proposed. Furthermore, additional priors and parameters for point selection, RANSAC and pose quality estimate (inlier test) are proposed and applied.