Multi-Features Guidance Network for partial-to-partial point cloud registration

TL;DR

This paper introduces a Multi-Features Guidance Network (MFG) for partial-to-partial point cloud registration that effectively combines shape and spatial features to improve matching accuracy and robustness.

Contribution

The proposed MFG network uniquely integrates shape features and spatial coordinates for correspondence search and credibility assessment, advancing partial-to-partial point cloud registration.

Findings

Outperforms current state-of-the-art methods in accuracy

Reduces impact of mismatched points through credibility scoring

Maintains high computational efficiency

Abstract

To eliminate the problems of large dimensional differences, big semantic gap, and mutual interference caused by hybrid features, in this paper, we propose a novel Multi-Features Guidance Network for partial-to-partial point cloud registration(MFG). The proposed network mainly includes four parts: keypoints' feature extraction, correspondences searching, correspondences credibility computation, and SVD, among which correspondences searching and correspondence credibility computation are the cores of the network. Unlike the previous work, we utilize the shape features and the spatial coordinates to guide correspondences search independently and fusing the matching results to obtain the final matching matrix. In the correspondences credibility computation module, based on the conflicted relationship between the features matching matrix and the coordinates matching matrix, we score the reliability for each correspondence, which can reduce the impact of mismatched or non-matched points. Experimental results show that our network outperforms the current state-of-the-art while maintaining computational efficiency.