Curb detection and accessibility evaluation from low-density mobile mapping point cloud data

Kiichiro Ishikawa*, Daisuke Kubo, Yoshiharu Amano

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)

Abstract

Our goal is to automatically classify objects from Mobile Mapping System data to enable the automatic construction of dynamic maps. We aimed at the extraction of curbstones and classification of curb types. Although there is much research about curbstones being recognized from laser-scanned point clouds, there are few methods to classify curb types. In this paper, we propose a method to extract curbstones from low-density-type laser scan data. We also propose a method to distinguish whether curbstones allow access to off-road facilities. Evaluation tests give an F-measure of ≥94.4% and an accessibility classification accuracy of ≥99.6%. Moreover, the results of applying multiple filters to noise removal are compared.

Original languageEnglish
Pages (from-to)376-385
Number of pages10
JournalInternational Journal of Automation Technology
Volume12
Issue number3
DOIs
Publication statusPublished - 2018 May

Keywords

  • Autonomous drive
  • Curb accessibility
  • Dynamic map
  • Mobile mapping system

ASJC Scopus subject areas

  • Mechanical Engineering
  • Industrial and Manufacturing Engineering

Fingerprint

Dive into the research topics of 'Curb detection and accessibility evaluation from low-density mobile mapping point cloud data'. Together they form a unique fingerprint.

Cite this