Benchmarking 6DOF Outdoor Visual Localization in Changing Conditions
Paper i proceeding, 2018

Visual localization enables autonomous vehicles to navigate in their surroundings and augmented reality applica-tions to link virtual to real worlds. Practical visual localization approaches need to be robust to a wide variety of viewing condition, including day-night changes, as well as weather and seasonal variations, while providing highly accurate 6 degree-of-freedom (6DOF) camera pose estimates.

In this paper, we introduce the first benchmark datasets specifically designed for analyzing the impact of such factors on visual localization. Using carefully created ground truth poses for query images taken under a wide variety of conditions, we evaluate the impact of various factors on 6DOF camera pose estimation accuracy through extensive experiments with state-of-the-art localization approaches.

Based on our results, we draw conclusions about the difficulty of different conditions, showing that long-term localization is far from solved, and propose promising avenues for future work, including sequence-based localization approaches and the need for better local features. Our benchmark is available at visuallocalization.net

Visual localization

camera pose estimation

long term localization

benchmark

Författare

Torsten Sattler

Eidgenössische Technische Hochschule Zürich (ETH)

Will Maddern

University of Oxford

Carl Toft

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik

Akihiko Torii

Tokyo Institute of Technology

Lars Hammarstrand

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik

Erik Stenborg

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik

Daniel Safari

Tokyo Institute of Technology

Danmarks Tekniske Universitet (DTU)

Masatoshi Okutomi

Tokyo Institute of Technology

Marc Pollefeys

Microsoft Corporation

Eidgenössische Technische Hochschule Zürich (ETH)

Josef Sivic

Ceske Vysoke Uceni Technicke v Praze

Institut National de Recherche en Informatique et en Automatique (INRIA)

Fredrik Kahl

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik

Lunds universitet

Tomas Pajdla

Ceske Vysoke Uceni Technicke v Praze

Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition

10636919 (ISSN)

8601-8610 8578995
978-153866420-9 (ISBN)

Conference on Computer Vision and Pattern Recognition 2018
Salt Lake City, USA,

COPPLAR CampusShuttle cooperative perception & planning platform

VINNOVA (2015-04849), 2016-01-01 -- 2018-12-31.

Styrkeområden

Informations- och kommunikationsteknik

Ämneskategorier

Robotteknik och automation

Sannolikhetsteori och statistik

Datorseende och robotik (autonoma system)

DOI

10.1109/CVPR.2018.00897

Mer information

Senast uppdaterat

2019-07-09