Semantic Match Consistency for Long-Term Visual Localization
Paper i proceeding, 2018

Robust and accurate visual localization across large appearance variations due to changes in time of day, seasons, or changes of the environment is a challenging problem which is of importance to application areas such as navigation of autonomous robots. Traditional feature-based methods often struggle in these conditions due to the significant number of erroneous matches between the image and the 3D model. In this paper, we present a method for scoring the individual correspondences by exploiting semantic information about the query image and the scene. In this way, erroneous correspondences tend to get a low semantic consistency score, whereas correct correspondences tend to get a high score. By incorporating this information in a standard localization pipeline, we show that the localization performance can be significantly improved compared to the state-of-the-art, as evaluated on two challenging long-term localization benchmarks.

Visual localization

Outlier rejection

Camera pose estimation

Semantic segmentation

Self-driving cars

Författare

Carl Toft

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik, Digitala bildsystem och bildanalys

Erik Stenborg

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik, Signalbehandling

Lars Hammarstrand

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik, Signalbehandling

Lucas Brynte

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik, Digitala bildsystem och bildanalys

Marc Pollefeys

Eidgenössische Technische Hochschule Zürich (ETH)

Microsoft

Torsten Sattler

Eidgenössische Technische Hochschule Zürich (ETH)

Fredrik Kahl

Chalmers, Elektroteknik, Signalbehandling och medicinsk teknik, Digitala bildsystem och bildanalys

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

03029743 (ISSN) 16113349 (eISSN)

Vol. 11206 LNCS 391-408

15th European Conference on Computer Vision, ECCV 2018
Munich, Germany,

COPPLAR CampusShuttle cooperative perception & planning platform

VINNOVA, 2016-01-01 -- 2018-12-31.

Ämneskategorier

Robotteknik och automation

Datavetenskap (datalogi)

Datorseende och robotik (autonoma system)

DOI

10.1007/978-3-030-01216-8_24

Mer information

Senast uppdaterat

2019-04-11