Geometric Multi-Session Map Merging with Learned Local Descriptors

TL;DR

This paper introduces GMLD, a learning-based local descriptor framework that enhances multi-session point cloud map merging by improving map alignment accuracy and robustness across different sessions and environments.

Contribution

The paper presents a novel learning-based local descriptor framework with a keypoint-aware encoder and geometric transformer for improved multi-session map merging.

Findings

Accurate and robust map merging demonstrated on public and self-collected datasets.

Learned features outperform traditional methods in loop closure detection.

Low error in global map alignment across diverse environments.

Abstract

Multi-session map merging is crucial for extended autonomous operations in large-scale environments. In this paper, we present GMLD, a learning-based local descriptor framework for large-scale multi-session point cloud map merging that systematically aligns maps collected across different sessions with overlapping regions. The proposed framework employs a keypoint-aware encoder and a plane-based geometric transformer to extract discriminative features for loop closure detection and relative pose estimation. To further improve global consistency, we include inter-session scan matching cost factors in the factor-graph optimization stage. We evaluate our framework on the public datasets, as well as self-collected data from diverse environments. The results show accurate and robust map merging with low error, and the learned features deliver strong performance in both loop closure detection and relative pose estimation.