Vehicle localization based on global navigation satellite system aided by three-dimensional map

Yanlei Gu, Li Ta Hsu, Shunsuke Kamijo

Research output: Chapter in book / Conference proceedingChapter in an edited book (as author)Academic researchpeer-review

1 Citation (Scopus)

Abstract

Accurate vehicle localization technologies are significant for current onboard navigation systems and future autonomous vehicles. More specifically, positioning accuracy is expected at the submeter level. This paper presents an accurate vehicle self-localization system and evaluates the proposed system in different classes of urban environments. The developed system adopts an innovative global navigation satellite system (GNSS) positioning method as the key technique. The GNSS positioning method can improve the positioning error by reducing the effects of multipath interference and non-line-of-sight errors with the aid of a three-dimensional map. To improve positioning accuracy further, the vehicle localization system integrates the GNSS positioning technique with inertial sensors and vision sensors by considering the characteristics of each sensor. The inertial sensors represent vehicle movement with heading direction and vehicle speed. The vision sensor is used to recognize the position change relative to lane markings on the road surface. Those techniques and sensors collaborate to provide an accurate position in the global coordinate system. To verify the effectiveness and stability of the proposed system, a series of tests was conducted in one of the most challenging urban cities, Tokyo. The experiment results demonstrate that the proposed system can achieve submeter accuracy for the positioning error mean and has a 90% correct lane rate in the localization.

Original languageEnglish
Title of host publicationIntelligent Transportation Systems
PublisherNational Research Council
Pages55-61
Number of pages7
ISBN (Electronic)9780309441964
DOIs
Publication statusPublished - 1 Jan 2017
Externally publishedYes

Publication series

NameTransportation Research Record
Volume2621
ISSN (Print)0361-1981

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Mechanical Engineering

Fingerprint

Dive into the research topics of 'Vehicle localization based on global navigation satellite system aided by three-dimensional map'. Together they form a unique fingerprint.

Cite this