An Approach to Robust ICP Initialization

Alexander Kolpakov; Michael Werman

doi:10.1109/TPAMI.2023.3287468

An Approach to Robust ICP Initialization

Alexander Kolpakov, Michael Werman

The Rachel and Selim Benin School of Engineering and Computer Science

The Hebrew University of Jerusalem

Research output: Contribution to journal › Article › peer-review

Abstract

In this note, we propose an approach to initialize the Iterative Closest Point (ICP) algorithm to match unlabelled point clouds related by rigid transformations. The method is based on matching the ellipsoids defined by the points' covariance matrices and then testing the various principal half-axes matchings that differ by elements of a finite reflection group. We derive bounds on the robustness of our approach to noise and numerical experiments confirm our theoretical findings.

Original language	English
Pages (from-to)	12685-12691
Number of pages	7
Journal	IEEE Transactions on Pattern Analysis and Machine Intelligence
Volume	45
Issue number	10
DOIs	https://doi.org/10.1109/TPAMI.2023.3287468
State	Published - 1 Oct 2023

Keywords

ICP
image registration
image stitching

All Science Journal Classification (ASJC) codes

Software
Computer Vision and Pattern Recognition
Computational Theory and Mathematics
Artificial Intelligence
Applied Mathematics

Access to Document

10.1109/TPAMI.2023.3287468

Cite this

@article{e0d4f64df8df41ff97b9252a41b6bc89,

title = "An Approach to Robust ICP Initialization",

abstract = "In this note, we propose an approach to initialize the Iterative Closest Point (ICP) algorithm to match unlabelled point clouds related by rigid transformations. The method is based on matching the ellipsoids defined by the points' covariance matrices and then testing the various principal half-axes matchings that differ by elements of a finite reflection group. We derive bounds on the robustness of our approach to noise and numerical experiments confirm our theoretical findings.",

keywords = "ICP, image registration, image stitching",

author = "Alexander Kolpakov and Michael Werman",

note = "Publisher Copyright: {\textcopyright} 1979-2012 IEEE.",

year = "2023",

month = oct,

day = "1",

doi = "10.1109/TPAMI.2023.3287468",

language = "الإنجليزيّة",

volume = "45",

pages = "12685--12691",

journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",

issn = "0162-8828",

publisher = "IEEE Computer Society",

number = "10",

}

TY - JOUR

T1 - An Approach to Robust ICP Initialization

AU - Kolpakov, Alexander

AU - Werman, Michael

PY - 2023/10/1

Y1 - 2023/10/1

N2 - In this note, we propose an approach to initialize the Iterative Closest Point (ICP) algorithm to match unlabelled point clouds related by rigid transformations. The method is based on matching the ellipsoids defined by the points' covariance matrices and then testing the various principal half-axes matchings that differ by elements of a finite reflection group. We derive bounds on the robustness of our approach to noise and numerical experiments confirm our theoretical findings.

AB - In this note, we propose an approach to initialize the Iterative Closest Point (ICP) algorithm to match unlabelled point clouds related by rigid transformations. The method is based on matching the ellipsoids defined by the points' covariance matrices and then testing the various principal half-axes matchings that differ by elements of a finite reflection group. We derive bounds on the robustness of our approach to noise and numerical experiments confirm our theoretical findings.

KW - ICP

KW - image registration

KW - image stitching

UR - http://www.scopus.com/inward/record.url?scp=85162910817&partnerID=8YFLogxK

U2 - 10.1109/TPAMI.2023.3287468

DO - 10.1109/TPAMI.2023.3287468

M3 - مقالة

C2 - 37335788

SN - 0162-8828

VL - 45

SP - 12685

EP - 12691

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

IS - 10

ER -

An Approach to Robust ICP Initialization

Abstract

Keywords

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this