SGNet: Salient Geometric Network for Point Cloud Registration

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F24%3A00380045" target="_blank" >RIV/68407700:21230/24:00380045 - isvavai.cz</a>

Result on the web

<a href="https://doi.org/10.1109/IROS58592.2024.10802262" target="_blank" >https://doi.org/10.1109/IROS58592.2024.10802262</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/IROS58592.2024.10802262" target="_blank" >10.1109/IROS58592.2024.10802262</a>

Result language

angličtina

Original language name

SGNet: Salient Geometric Network for Point Cloud Registration

Original language description

Point Cloud Registration (PCR) is a critical and challenging task in computer vision and robotics. One of the primary diffculties in PCR is identifying salient and meaningful points that exhibit consistent semantic and geometric properties across different scans. Previous methods have encountered challenges with ambiguous matching due to the similarity among patch blocks throughout the entire point cloud and the lack of consideration for effcient global geometric consistency. To address these issues, we propose a new framework that includes several novel techniques. Firstly, we introduce a semantic-aware geometric encoder that combines object-level and patch-level semantic information. This encoder signifcantly improves registration recall by reducing ambiguity in patchlevel superpoint matching. Additionally, we incorporate a prior knowledge approach that utilizes an intrinsic shape signature to identify salient points. This enables us to extract the most salient super points and meaningful dense points in the scene. Secondly, we introduce an innovative transformer that encodes High-Order (HO) geometric features. These features are crucial for identifying salient points within initial overlap regions while considering global high-order geometric consistency. We introduce an anchor node selection strategy to optimize this highorder transformer further. By encoding inter-frame triangle or polyhedron consistency features based on these anchor nodes, we can effectively learn high-order geometric features of salient super points. These high-order features are then propagated to dense points and utilized by a Sinkhorn matching module to identify critical correspondences for successful registration. The experiments conducted on the 3DMatch/3DLoMatch and KITTI datasets demonstrate the effectiveness of our method.

Czech name

—

Czech description

—

Type

D - Article in proceedings

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

<a href="/en/project/GM22-23183M" target="_blank" >GM22-23183M: New generation of camera geometry solvers</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year

2024

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024)

ISBN

979-8-3503-7771-2

ISSN

2153-0858

e-ISSN

2153-0866

Number of pages

7

Pages from-to

3276-3282

Publisher name

IEEE

Place of publication

Piscataway

Event location

Abu Dhabi

Event date

Oct 14, 2024

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

001411890000353