A maximum feasible subsystem for globally optimal 3D point cloud registration

Chanki Yu, Da Young Ju

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In this paper, a globally optimal algorithm based on a maximum feasible subsystem framework is proposed for robust pairwise registration of point cloud data. Registration is formulated as a branch-and-bound problem with mixed-integer linear programming. Among the putative matches of three-dimensional (3D) features between two sets of range data, the proposed algorithm finds the maximum number of geometrically correct correspondences in the presence of incorrect matches, and it estimates the transformation parameters in a globally optimal manner. The optimization requires no initialization of transformation parameters. Experimental results demonstrated that the presented algorithm was more accurate and reliable than state-of-the-art registration methods and showed robustness against severe outliers/mismatches. This global optimization technique was highly effective, even when the geometric overlap between the datasets was very small.

Original languageEnglish
Article number544
JournalSensors (Switzerland)
Volume18
Issue number2
DOIs
Publication statusPublished - 2018 Feb 10

Fingerprint

Linear Programming
linear programming
optimization
Global optimization
Linear programming
integers
estimates
Datasets

All Science Journal Classification (ASJC) codes

  • Analytical Chemistry
  • Atomic and Molecular Physics, and Optics
  • Biochemistry
  • Instrumentation
  • Electrical and Electronic Engineering

Cite this

@article{37968d836a244ca1837594ceffa154da,
title = "A maximum feasible subsystem for globally optimal 3D point cloud registration",
abstract = "In this paper, a globally optimal algorithm based on a maximum feasible subsystem framework is proposed for robust pairwise registration of point cloud data. Registration is formulated as a branch-and-bound problem with mixed-integer linear programming. Among the putative matches of three-dimensional (3D) features between two sets of range data, the proposed algorithm finds the maximum number of geometrically correct correspondences in the presence of incorrect matches, and it estimates the transformation parameters in a globally optimal manner. The optimization requires no initialization of transformation parameters. Experimental results demonstrated that the presented algorithm was more accurate and reliable than state-of-the-art registration methods and showed robustness against severe outliers/mismatches. This global optimization technique was highly effective, even when the geometric overlap between the datasets was very small.",
author = "Chanki Yu and Ju, {Da Young}",
year = "2018",
month = "2",
day = "10",
doi = "10.3390/s18020544",
language = "English",
volume = "18",
journal = "Sensors",
issn = "1424-3210",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "2",

}

A maximum feasible subsystem for globally optimal 3D point cloud registration. / Yu, Chanki; Ju, Da Young.

In: Sensors (Switzerland), Vol. 18, No. 2, 544, 10.02.2018.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A maximum feasible subsystem for globally optimal 3D point cloud registration

AU - Yu, Chanki

AU - Ju, Da Young

PY - 2018/2/10

Y1 - 2018/2/10

N2 - In this paper, a globally optimal algorithm based on a maximum feasible subsystem framework is proposed for robust pairwise registration of point cloud data. Registration is formulated as a branch-and-bound problem with mixed-integer linear programming. Among the putative matches of three-dimensional (3D) features between two sets of range data, the proposed algorithm finds the maximum number of geometrically correct correspondences in the presence of incorrect matches, and it estimates the transformation parameters in a globally optimal manner. The optimization requires no initialization of transformation parameters. Experimental results demonstrated that the presented algorithm was more accurate and reliable than state-of-the-art registration methods and showed robustness against severe outliers/mismatches. This global optimization technique was highly effective, even when the geometric overlap between the datasets was very small.

AB - In this paper, a globally optimal algorithm based on a maximum feasible subsystem framework is proposed for robust pairwise registration of point cloud data. Registration is formulated as a branch-and-bound problem with mixed-integer linear programming. Among the putative matches of three-dimensional (3D) features between two sets of range data, the proposed algorithm finds the maximum number of geometrically correct correspondences in the presence of incorrect matches, and it estimates the transformation parameters in a globally optimal manner. The optimization requires no initialization of transformation parameters. Experimental results demonstrated that the presented algorithm was more accurate and reliable than state-of-the-art registration methods and showed robustness against severe outliers/mismatches. This global optimization technique was highly effective, even when the geometric overlap between the datasets was very small.

UR - http://www.scopus.com/inward/record.url?scp=85041944418&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85041944418&partnerID=8YFLogxK

U2 - 10.3390/s18020544

DO - 10.3390/s18020544

M3 - Article

AN - SCOPUS:85041944418

VL - 18

JO - Sensors

JF - Sensors

SN - 1424-3210

IS - 2

M1 - 544

ER -