Journal of Geographical Systems

, Volume 18, Issue 3, pp 249–264

Efficient road geometry identification from digital vector data

Original Article

DOI: 10.1007/s10109-016-0230-1

Cite this article as:
Andrášik, R. & Bíl, M. J Geogr Syst (2016) 18: 249. doi:10.1007/s10109-016-0230-1


A new method for the automatic identification of road geometry from digital vector data is presented. The method is capable of efficiently identifying circular curves with their radii and tangents (straight sections). The average error of identification ranged from 0.01 to 1.30 % for precisely drawn data and 4.81 % in the case of actual road data with noise in the location of vertices. The results demonstrate that the proposed method is faster and more precise than commonly used techniques. This approach can be used by road administrators to complete their databases with information concerning the geometry of roads. It can also be utilized by transport engineers or traffic safety analysts to investigate the possible dependence of traffic accidents on road geometries. The method presented is applicable as well to railroads and rivers or other line features.


Circular curves Tangents Automatic geometry identification Curvature Discriminant analysis Classification tree Roads Database GIS 

JEL Classification

C8 C18 R41 

Funding information

Funder NameGrant NumberFunding Note
Ministry of Education, Youth and Sports (CZ)
  • CZ.1.05/2.1.00/03.0064
Ministerstvo vnitra
  • VI20152019049

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  1. 1.Department of GeoinformaticsCDV Transport Research CentreBrnoCzech Republic

Personalised recommendations