A fully automatic approach to register mobile mapping and airborne imagery to support the correction of platform trajectories in GNSS-denied urban areas

P.L.H. Jende (Corresponding Author), F.C. Nex, Markus Gerke, G. Vosselman

Research output: Contribution to journalArticleAcademicpeer-review

6 Citations (Scopus)

Abstract

Mobile Mapping (MM) solutions have become a significant extension to traditional data acquisition methods over the last years. Independently from the sensor carried by a platform, may it be laser scanners or cameras, high-resolution data postings are opposing a poor absolute localisation accuracy in urban areas due to GNSS occlusions and multipath effects. Potentially inaccurate position estimations are propagated by IMUs which are furthermore prone to drift effects.

Thus, reliable and accurate absolute positioning on a par with MM’s high-quality data remains an open issue. Multiple and diverse approaches have shown promising potential to mitigate GNSS errors in urban areas, but cannot achieve decimetre accuracy, require manual effort, or have limitations with respect to costs and availability.

This paper presents a fully automatic approach to support the correction of MM imaging data based on correspondences with airborne nadir images. These correspondences can be employed to correct the MM platform’s orientation by an adjustment solution. Unlike MM as such, aerial images do not suffer from GNSS occlusions, and their accuracy is usually verified by employing well-established methods using ground control points. However, a registration between MM and aerial images is a non-standard matching scenario, and requires several strategies to yield reliable and accurate correspondences. Scale, perspective and content strongly vary between both image sources, thus traditional feature matching methods may fail. To this end, the registration process is designed to focus on common and clearly distinguishable elements, such as road markings, manholes, or kerbstones. With a registration accuracy of about 98%, reliable tie information between MM and aerial data can be derived. Even though, the adjustment strategy is not covered in its entirety in this paper, accuracy results after adjustment will be presented. It will be shown that a decimetre accuracy is well achievable in a real data test scenario.
Original languageEnglish
Pages (from-to)86-99
Number of pages14
JournalISPRS journal of photogrammetry and remote sensing
Volume141
DOIs
Publication statusPublished - 1 Jul 2018

Fingerprint

airborne sensing
GNSS
registers
imagery
platforms
urban area
trajectory
Trajectories
trajectories
occlusion
adjusting
Antennas
ground control
nadir
scanner
roads
data quality
data acquisition
positioning
scanners

Keywords

  • ITC-ISI-JOURNAL-ARTICLE

Cite this

@article{2e2829372e2949569ca24b9da7c06767,
title = "A fully automatic approach to register mobile mapping and airborne imagery to support the correction of platform trajectories in GNSS-denied urban areas",
abstract = "Mobile Mapping (MM) solutions have become a significant extension to traditional data acquisition methods over the last years. Independently from the sensor carried by a platform, may it be laser scanners or cameras, high-resolution data postings are opposing a poor absolute localisation accuracy in urban areas due to GNSS occlusions and multipath effects. Potentially inaccurate position estimations are propagated by IMUs which are furthermore prone to drift effects.Thus, reliable and accurate absolute positioning on a par with MM’s high-quality data remains an open issue. Multiple and diverse approaches have shown promising potential to mitigate GNSS errors in urban areas, but cannot achieve decimetre accuracy, require manual effort, or have limitations with respect to costs and availability.This paper presents a fully automatic approach to support the correction of MM imaging data based on correspondences with airborne nadir images. These correspondences can be employed to correct the MM platform’s orientation by an adjustment solution. Unlike MM as such, aerial images do not suffer from GNSS occlusions, and their accuracy is usually verified by employing well-established methods using ground control points. However, a registration between MM and aerial images is a non-standard matching scenario, and requires several strategies to yield reliable and accurate correspondences. Scale, perspective and content strongly vary between both image sources, thus traditional feature matching methods may fail. To this end, the registration process is designed to focus on common and clearly distinguishable elements, such as road markings, manholes, or kerbstones. With a registration accuracy of about 98{\%}, reliable tie information between MM and aerial data can be derived. Even though, the adjustment strategy is not covered in its entirety in this paper, accuracy results after adjustment will be presented. It will be shown that a decimetre accuracy is well achievable in a real data test scenario.",
keywords = "ITC-ISI-JOURNAL-ARTICLE",
author = "P.L.H. Jende and F.C. Nex and Markus Gerke and G. Vosselman",
year = "2018",
month = "7",
day = "1",
doi = "10.1016/j.isprsjprs.2018.04.017",
language = "English",
volume = "141",
pages = "86--99",
journal = "ISPRS journal of photogrammetry and remote sensing",
issn = "0924-2716",
publisher = "Elsevier",

}

A fully automatic approach to register mobile mapping and airborne imagery to support the correction of platform trajectories in GNSS-denied urban areas. / Jende, P.L.H. (Corresponding Author); Nex, F.C.; Gerke, Markus; Vosselman, G.

In: ISPRS journal of photogrammetry and remote sensing, Vol. 141, 01.07.2018, p. 86-99.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - A fully automatic approach to register mobile mapping and airborne imagery to support the correction of platform trajectories in GNSS-denied urban areas

AU - Jende, P.L.H.

AU - Nex, F.C.

AU - Gerke, Markus

AU - Vosselman, G.

PY - 2018/7/1

Y1 - 2018/7/1

N2 - Mobile Mapping (MM) solutions have become a significant extension to traditional data acquisition methods over the last years. Independently from the sensor carried by a platform, may it be laser scanners or cameras, high-resolution data postings are opposing a poor absolute localisation accuracy in urban areas due to GNSS occlusions and multipath effects. Potentially inaccurate position estimations are propagated by IMUs which are furthermore prone to drift effects.Thus, reliable and accurate absolute positioning on a par with MM’s high-quality data remains an open issue. Multiple and diverse approaches have shown promising potential to mitigate GNSS errors in urban areas, but cannot achieve decimetre accuracy, require manual effort, or have limitations with respect to costs and availability.This paper presents a fully automatic approach to support the correction of MM imaging data based on correspondences with airborne nadir images. These correspondences can be employed to correct the MM platform’s orientation by an adjustment solution. Unlike MM as such, aerial images do not suffer from GNSS occlusions, and their accuracy is usually verified by employing well-established methods using ground control points. However, a registration between MM and aerial images is a non-standard matching scenario, and requires several strategies to yield reliable and accurate correspondences. Scale, perspective and content strongly vary between both image sources, thus traditional feature matching methods may fail. To this end, the registration process is designed to focus on common and clearly distinguishable elements, such as road markings, manholes, or kerbstones. With a registration accuracy of about 98%, reliable tie information between MM and aerial data can be derived. Even though, the adjustment strategy is not covered in its entirety in this paper, accuracy results after adjustment will be presented. It will be shown that a decimetre accuracy is well achievable in a real data test scenario.

AB - Mobile Mapping (MM) solutions have become a significant extension to traditional data acquisition methods over the last years. Independently from the sensor carried by a platform, may it be laser scanners or cameras, high-resolution data postings are opposing a poor absolute localisation accuracy in urban areas due to GNSS occlusions and multipath effects. Potentially inaccurate position estimations are propagated by IMUs which are furthermore prone to drift effects.Thus, reliable and accurate absolute positioning on a par with MM’s high-quality data remains an open issue. Multiple and diverse approaches have shown promising potential to mitigate GNSS errors in urban areas, but cannot achieve decimetre accuracy, require manual effort, or have limitations with respect to costs and availability.This paper presents a fully automatic approach to support the correction of MM imaging data based on correspondences with airborne nadir images. These correspondences can be employed to correct the MM platform’s orientation by an adjustment solution. Unlike MM as such, aerial images do not suffer from GNSS occlusions, and their accuracy is usually verified by employing well-established methods using ground control points. However, a registration between MM and aerial images is a non-standard matching scenario, and requires several strategies to yield reliable and accurate correspondences. Scale, perspective and content strongly vary between both image sources, thus traditional feature matching methods may fail. To this end, the registration process is designed to focus on common and clearly distinguishable elements, such as road markings, manholes, or kerbstones. With a registration accuracy of about 98%, reliable tie information between MM and aerial data can be derived. Even though, the adjustment strategy is not covered in its entirety in this paper, accuracy results after adjustment will be presented. It will be shown that a decimetre accuracy is well achievable in a real data test scenario.

KW - ITC-ISI-JOURNAL-ARTICLE

UR - https://ezproxy2.utwente.nl/login?url=https://webapps.itc.utwente.nl/library/2018/isi/jende_ful.pdf

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1016/j.isprsjprs.2018.04.017

U2 - 10.1016/j.isprsjprs.2018.04.017

DO - 10.1016/j.isprsjprs.2018.04.017

M3 - Article

VL - 141

SP - 86

EP - 99

JO - ISPRS journal of photogrammetry and remote sensing

JF - ISPRS journal of photogrammetry and remote sensing

SN - 0924-2716

ER -