Contextual segment-based classification of airborne laser scanner data

George Vosselman*, Maximilian Coenen, Franz Rottensteiner

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

41 Citations (Scopus)
2 Downloads (Pure)

Abstract

Classification of point clouds is needed as a first step in the extraction of various types of geo-information from point clouds. We present a new approach to contextual classification of segmented airborne laser scanning data. Potential advantages of segment-based classification are easily offset by segmentation errors. We combine different point cloud segmentation methods to minimise both under- and over-segmentation. We propose a contextual segment-based classification using a Conditional Random Field. Segment adjacencies are represented by edges in the graphical model and characterised by a range of features of points along the segment borders. A mix of small and large segments allows the interaction between nearby and distant points. Results of the segment-based classification are compared to results of a point-based CRF classification. Whereas only a small advantage of the segment-based classification is observed for the ISPRS Vaihingen dataset with 4–7 points/m2, the percentage of correctly classified points in a 30 points/m2 dataset of Rotterdam amounts to 91.0% for the segment-based classification vs. 82.8% for the point-based classification.

Original languageEnglish
Pages (from-to)354-371
Number of pages18
JournalISPRS journal of photogrammetry and remote sensing
Volume128
DOIs
Publication statusPublished - 1 Jun 2017

Fingerprint

airborne lasers
scanner
scanners
laser
Lasers
segmentation
borders
Scanning

Keywords

  • Classification
  • CRF
  • Point cloud
  • Segmentation
  • ITC-ISI-JOURNAL-ARTICLE

Cite this

@article{a7a03fd92c42485cbb643726a55879de,
title = "Contextual segment-based classification of airborne laser scanner data",
abstract = "Classification of point clouds is needed as a first step in the extraction of various types of geo-information from point clouds. We present a new approach to contextual classification of segmented airborne laser scanning data. Potential advantages of segment-based classification are easily offset by segmentation errors. We combine different point cloud segmentation methods to minimise both under- and over-segmentation. We propose a contextual segment-based classification using a Conditional Random Field. Segment adjacencies are represented by edges in the graphical model and characterised by a range of features of points along the segment borders. A mix of small and large segments allows the interaction between nearby and distant points. Results of the segment-based classification are compared to results of a point-based CRF classification. Whereas only a small advantage of the segment-based classification is observed for the ISPRS Vaihingen dataset with 4–7 points/m2, the percentage of correctly classified points in a 30 points/m2 dataset of Rotterdam amounts to 91.0{\%} for the segment-based classification vs. 82.8{\%} for the point-based classification.",
keywords = "Classification, CRF, Point cloud, Segmentation, ITC-ISI-JOURNAL-ARTICLE",
author = "George Vosselman and Maximilian Coenen and Franz Rottensteiner",
year = "2017",
month = "6",
day = "1",
doi = "10.1016/j.isprsjprs.2017.03.010",
language = "English",
volume = "128",
pages = "354--371",
journal = "ISPRS journal of photogrammetry and remote sensing",
issn = "0924-2716",
publisher = "Elsevier",

}

Contextual segment-based classification of airborne laser scanner data. / Vosselman, George; Coenen, Maximilian; Rottensteiner, Franz.

In: ISPRS journal of photogrammetry and remote sensing, Vol. 128, 01.06.2017, p. 354-371.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Contextual segment-based classification of airborne laser scanner data

AU - Vosselman, George

AU - Coenen, Maximilian

AU - Rottensteiner, Franz

PY - 2017/6/1

Y1 - 2017/6/1

N2 - Classification of point clouds is needed as a first step in the extraction of various types of geo-information from point clouds. We present a new approach to contextual classification of segmented airborne laser scanning data. Potential advantages of segment-based classification are easily offset by segmentation errors. We combine different point cloud segmentation methods to minimise both under- and over-segmentation. We propose a contextual segment-based classification using a Conditional Random Field. Segment adjacencies are represented by edges in the graphical model and characterised by a range of features of points along the segment borders. A mix of small and large segments allows the interaction between nearby and distant points. Results of the segment-based classification are compared to results of a point-based CRF classification. Whereas only a small advantage of the segment-based classification is observed for the ISPRS Vaihingen dataset with 4–7 points/m2, the percentage of correctly classified points in a 30 points/m2 dataset of Rotterdam amounts to 91.0% for the segment-based classification vs. 82.8% for the point-based classification.

AB - Classification of point clouds is needed as a first step in the extraction of various types of geo-information from point clouds. We present a new approach to contextual classification of segmented airborne laser scanning data. Potential advantages of segment-based classification are easily offset by segmentation errors. We combine different point cloud segmentation methods to minimise both under- and over-segmentation. We propose a contextual segment-based classification using a Conditional Random Field. Segment adjacencies are represented by edges in the graphical model and characterised by a range of features of points along the segment borders. A mix of small and large segments allows the interaction between nearby and distant points. Results of the segment-based classification are compared to results of a point-based CRF classification. Whereas only a small advantage of the segment-based classification is observed for the ISPRS Vaihingen dataset with 4–7 points/m2, the percentage of correctly classified points in a 30 points/m2 dataset of Rotterdam amounts to 91.0% for the segment-based classification vs. 82.8% for the point-based classification.

KW - Classification

KW - CRF

KW - Point cloud

KW - Segmentation

KW - ITC-ISI-JOURNAL-ARTICLE

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1016/j.isprsjprs.2017.03.010

UR - https://ezproxy2.utwente.nl/login?url=https://webapps.itc.utwente.nl/library/2017/isi/vosselman_con.pdf

U2 - 10.1016/j.isprsjprs.2017.03.010

DO - 10.1016/j.isprsjprs.2017.03.010

M3 - Article

AN - SCOPUS:85018625462

VL - 128

SP - 354

EP - 371

JO - ISPRS journal of photogrammetry and remote sensing

JF - ISPRS journal of photogrammetry and remote sensing

SN - 0924-2716

ER -