Abstract:The increasing availability of point cloud acquisition techniques makes it possible to significantly increase 3D observation capacity by the registration of multi-sensor, multi-platform, and multi-temporal point clouds. However, there are geometric heterogeneities (point density variations and point distribution differences), small overlaps (30 % ∼ 50 %), and large data amounts (a few millions) among these large-scale heterogeneous point clouds, which pose great challenges for effective and efficient registration. In this paper, considering the structural representation capacity of graph model, we propose an incremental registration method for large-scale heterogeneous point clouds by hierarchical graph matching. More specifically, we first construct a novel graph model to discriminatively and robustly represent heterogeneous point clouds. In addition to conventional nodes and edges, our graph model particularly designs discriminative and robust feature descriptors for local node description and captures spatial relationships from both locations and orientations for global edge description. We further devise a matching strategy to accurately estimate node matches for our graph models with partial even small overlaps. This effectiveness benefits from the comprehensiveness of node and edge dissimilarities and the constraint of geometric consistency in the optimization objective. On this basis, we design a coarse-to-fine registration framework for effective and efficient point cloud registration. In this incremental framework, graph matching is hierarchically utilized to achieve sparse-to-dense point matching by global extraction and local propagation, which provides dense correspondences for robust coarse registration and predicts overlap ratio for accurate fine registration, and also avoids huge computation costs for large-scale point clouds. Extensive experiments on one benchmark and three changing self-built datasets with large scales, outliers, changing densities, and small overlaps show the excellent transformation and correspondence accuracies of our registration method for large-scale heterogeneous point clouds. Compared to the state-of-the-art methods ( i.e., TrimICP, CoBigICP, GROR, VPFBR, DPCR, and PRR), our registration method performs approximate even higher efficiency while achieves an improvement of 33 % − 88 % regarding registration accuracy (OE).

Accelerating Point Cloud Registration with Low Overlap using Graphs and Sparse Convolutions

A Two-Stage Correspondence-Free Algorithm for Partially Overlapping Point Cloud Registration

Efficient low-overlapping point cloud registration based on two-stage network learning

Incremental registration towards large-scale heterogeneous point clouds by hierarchical graph matching

Low-Overlap Point Cloud Registration With Transformer

Iterative Overlap Attention-Aware Network with Similarity Learning for Partial Point Cloud Registration

Sparse and Low-Overlapping Point Cloud Registration Network for Indoor Building Environments

Point Cloud Registration using Representative Overlapping Points

Robust Point Cloud Registration Framework Based on Deep Graph Matching

Graph Matching Optimization Network for Point Cloud Registration

Point Cloud Registration Based on Graph Matching Optimization

Semantic enhancement based adaptive geometric encoding network for low overlap point cloud registration

Consistent Two-Flow Network for Tele-Registration of Point Clouds

Fast Robust Point Cloud Registration Based on Compatibility Graph and Accelerated Guided Sampling

Low Overlapping Point Cloud Registration using Mutual Prior based Completion Network

A fast point cloud registration method based on spatial relations and features

Geometric Transformer for Fast and Robust Point Cloud Registration

Geometry-Aware Enhanced Mutual-Supervised Point Elimination with Overlapping Mask Contrastive Learning for Partitial Point Cloud Registration

Deep learning-based low overlap point cloud registration for complex scenario: The review

CFI2P: Coarse-to-Fine Cross-Modal Correspondence Learning for Image-to-Point Cloud Registration

SACF-Net: Skip-attention Based Correspondence Filtering Network for Point Cloud Registration