Alignment of point cloud data acquired from continuous view points on flat surface

Kenta Ochiai, Masafumi Nakagawa

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A three-dimensional point cloud data measured with terrestrial 3D scanner is suitable for a spatial representation in a plant management, disaster monitoring and verification in traffic accident. Moreover, the latest 3D scanners can acquire massive point cloud data in wide range for a short time. However, 3D data measured in outdoor environment contains optical errors caused by a mirror reflection and point noises caused by movers. Therefore, generally, an additional 3D measurement is conducted and the additional data are integrated to an initial measured point cloud data with a procedure of 3D data alignment. Moreover, conventional 3D alignment methodology requires geometrical features. In other words, these approaches are limited to uneven surface. Therefore, a flat surface is a difficult object for the conventional data alignment methodology. In our experiment, Time-of-Flight camera was used as a handheld 3D scanner. Moreover, we used infrared images taken from the camera as feature values. Then, we conducted an alignment of point cloud data acquired from continuous view points on flat surfaces. Additionally, we have confirmed that our approach can integrate point cloud data even if measured object is a flat surface.

Original languageEnglish
Title of host publication33rd Asian Conference on Remote Sensing 2012, ACRS 2012
Pages1666-1671
Number of pages6
Volume2
Publication statusPublished - 2012
Event33rd Asian Conference on Remote Sensing 2012, ACRS 2012 - Pattaya
Duration: 2012 Nov 262012 Nov 30

Other

Other33rd Asian Conference on Remote Sensing 2012, ACRS 2012
CityPattaya
Period12/11/2612/11/30

Fingerprint

Cameras
Plant management
Highway accidents
Disasters
Mirrors
Infrared radiation
Monitoring
Experiments

Keywords

  • Flat surface
  • Handheld 3D scanner
  • Iterative Closest Point (ICP)
  • Simultaneous Localization and Mapping (SLAM)

ASJC Scopus subject areas

  • Information Systems

Cite this

Ochiai, K., & Nakagawa, M. (2012). Alignment of point cloud data acquired from continuous view points on flat surface. In 33rd Asian Conference on Remote Sensing 2012, ACRS 2012 (Vol. 2, pp. 1666-1671)

Alignment of point cloud data acquired from continuous view points on flat surface. / Ochiai, Kenta; Nakagawa, Masafumi.

33rd Asian Conference on Remote Sensing 2012, ACRS 2012. Vol. 2 2012. p. 1666-1671.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ochiai, K & Nakagawa, M 2012, Alignment of point cloud data acquired from continuous view points on flat surface. in 33rd Asian Conference on Remote Sensing 2012, ACRS 2012. vol. 2, pp. 1666-1671, 33rd Asian Conference on Remote Sensing 2012, ACRS 2012, Pattaya, 12/11/26.
Ochiai K, Nakagawa M. Alignment of point cloud data acquired from continuous view points on flat surface. In 33rd Asian Conference on Remote Sensing 2012, ACRS 2012. Vol. 2. 2012. p. 1666-1671
Ochiai, Kenta ; Nakagawa, Masafumi. / Alignment of point cloud data acquired from continuous view points on flat surface. 33rd Asian Conference on Remote Sensing 2012, ACRS 2012. Vol. 2 2012. pp. 1666-1671
@inproceedings{280566465edd445cb77c013458035664,
title = "Alignment of point cloud data acquired from continuous view points on flat surface",
abstract = "A three-dimensional point cloud data measured with terrestrial 3D scanner is suitable for a spatial representation in a plant management, disaster monitoring and verification in traffic accident. Moreover, the latest 3D scanners can acquire massive point cloud data in wide range for a short time. However, 3D data measured in outdoor environment contains optical errors caused by a mirror reflection and point noises caused by movers. Therefore, generally, an additional 3D measurement is conducted and the additional data are integrated to an initial measured point cloud data with a procedure of 3D data alignment. Moreover, conventional 3D alignment methodology requires geometrical features. In other words, these approaches are limited to uneven surface. Therefore, a flat surface is a difficult object for the conventional data alignment methodology. In our experiment, Time-of-Flight camera was used as a handheld 3D scanner. Moreover, we used infrared images taken from the camera as feature values. Then, we conducted an alignment of point cloud data acquired from continuous view points on flat surfaces. Additionally, we have confirmed that our approach can integrate point cloud data even if measured object is a flat surface.",
keywords = "Flat surface, Handheld 3D scanner, Iterative Closest Point (ICP), Simultaneous Localization and Mapping (SLAM)",
author = "Kenta Ochiai and Masafumi Nakagawa",
year = "2012",
language = "English",
isbn = "9781622769742",
volume = "2",
pages = "1666--1671",
booktitle = "33rd Asian Conference on Remote Sensing 2012, ACRS 2012",

}

TY - GEN

T1 - Alignment of point cloud data acquired from continuous view points on flat surface

AU - Ochiai, Kenta

AU - Nakagawa, Masafumi

PY - 2012

Y1 - 2012

N2 - A three-dimensional point cloud data measured with terrestrial 3D scanner is suitable for a spatial representation in a plant management, disaster monitoring and verification in traffic accident. Moreover, the latest 3D scanners can acquire massive point cloud data in wide range for a short time. However, 3D data measured in outdoor environment contains optical errors caused by a mirror reflection and point noises caused by movers. Therefore, generally, an additional 3D measurement is conducted and the additional data are integrated to an initial measured point cloud data with a procedure of 3D data alignment. Moreover, conventional 3D alignment methodology requires geometrical features. In other words, these approaches are limited to uneven surface. Therefore, a flat surface is a difficult object for the conventional data alignment methodology. In our experiment, Time-of-Flight camera was used as a handheld 3D scanner. Moreover, we used infrared images taken from the camera as feature values. Then, we conducted an alignment of point cloud data acquired from continuous view points on flat surfaces. Additionally, we have confirmed that our approach can integrate point cloud data even if measured object is a flat surface.

AB - A three-dimensional point cloud data measured with terrestrial 3D scanner is suitable for a spatial representation in a plant management, disaster monitoring and verification in traffic accident. Moreover, the latest 3D scanners can acquire massive point cloud data in wide range for a short time. However, 3D data measured in outdoor environment contains optical errors caused by a mirror reflection and point noises caused by movers. Therefore, generally, an additional 3D measurement is conducted and the additional data are integrated to an initial measured point cloud data with a procedure of 3D data alignment. Moreover, conventional 3D alignment methodology requires geometrical features. In other words, these approaches are limited to uneven surface. Therefore, a flat surface is a difficult object for the conventional data alignment methodology. In our experiment, Time-of-Flight camera was used as a handheld 3D scanner. Moreover, we used infrared images taken from the camera as feature values. Then, we conducted an alignment of point cloud data acquired from continuous view points on flat surfaces. Additionally, we have confirmed that our approach can integrate point cloud data even if measured object is a flat surface.

KW - Flat surface

KW - Handheld 3D scanner

KW - Iterative Closest Point (ICP)

KW - Simultaneous Localization and Mapping (SLAM)

UR - http://www.scopus.com/inward/record.url?scp=84880011651&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84880011651&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781622769742

VL - 2

SP - 1666

EP - 1671

BT - 33rd Asian Conference on Remote Sensing 2012, ACRS 2012

ER -