Intelligent Viaduct Recognition and Driving Altitude Determination Using GPS Data

Li Ta Hsu, Yanlei Gu, Shunsuke Kamijo

Research output: Journal article publicationJournal articleAcademic researchpeer-review

11 Citations (Scopus)


The rapid development of road networks in highly urbanized cities requires a substantial number of viaducts to reduce the increasing traffic burden on urban highways. Unfortunately, this road design deteriorates the performance of global positioning system (GPS) navigators due to the signal blockage between satellite and receiver. As a result, it is difficult for GPS navigators to determine whether the vehicle is driving on the street or viaduct. Misleading information could confuse drivers and lead them to drive irregularly, which is dangerous in heavy traffic. This paper proposes a novel classification algorithm based on the fact that different satellite signals and conditions can be observed in the on-street and on-viaduct cases by the implementation of dynamic Bayesian network (DBN) to distinguish the driving area of a vehicle. In addition, the proposed DBN can also accurately estimate the driving altitude of the vehicle according to the experiment results.

Original languageEnglish
Article number8003480
Pages (from-to)175-184
Number of pages10
JournalIEEE Transactions on Intelligent Vehicles
Issue number3
Publication statusPublished - Sept 2017


  • Autonomous driving
  • GNSS
  • GPS
  • land application
  • localization
  • navigation
  • urban canyon

ASJC Scopus subject areas

  • Artificial Intelligence
  • Automotive Engineering
  • Control and Optimization


Dive into the research topics of 'Intelligent Viaduct Recognition and Driving Altitude Determination Using GPS Data'. Together they form a unique fingerprint.

Cite this