On scale initialization in non-overlapping multi-perspective visual odometry

Yifu Wang*, Laurent Kneip

*Corresponding author for this work

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    3 Citations (Scopus)

    Abstract

    Multi-perspective camera systems pointing into all directions represent an increasingly interesting solution for visual localization and mapping. They combine the benefits of omni-directional measurements with a sufficient baseline for producing measurements in metric scale. However, the observability of metric scale suffers from degenerate cases if the cameras do not share any overlap in their field of view. This problem is of particular importance in many relevant practical applications, and it impacts most heavily on the difficulty of bootstrapping the structure-from-motion process. The present paper introduces a complete real-time pipeline for visual odometry with non-overlapping, multi-perspective camera systems, and in particular presents a solution to the scale initialization problem. We evaluate our method on both simulated and real data, thus proving robust initialization capacity as well as best-in-class performance regarding the overall motion estimation accuracy.

    Original languageEnglish
    Title of host publicationComputer Vision Systems - 11th International Conference, ICVS 2017, Revised Selected Papers
    EditorsMarkus Vincze, Haoyao Chen, Ming Liu
    PublisherSpringer Verlag
    Pages144-157
    Number of pages14
    ISBN (Print)9783319683447
    DOIs
    Publication statusPublished - 2017
    Event11th International Conference on Computer Vision Systems, ICVS 2017 - Shenzhen, China
    Duration: 10 Jul 201713 Jul 2017

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume10528 LNCS
    ISSN (Print)0302-9743
    ISSN (Electronic)1611-3349

    Conference

    Conference11th International Conference on Computer Vision Systems, ICVS 2017
    Country/TerritoryChina
    CityShenzhen
    Period10/07/1713/07/17

    Fingerprint

    Dive into the research topics of 'On scale initialization in non-overlapping multi-perspective visual odometry'. Together they form a unique fingerprint.

    Cite this