Pole-like roadside objects extraction from mobile lidar point clouds

Chi Min Chiu, Tee-Ann Teo*, Yao Tsung Lin, Shin Hui Li

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Mobile lidar system can acquire very dense and accurate 3-D point clouds along the road corridors. The acquiring data can provide plentiful road information such as road surface, road marks and road objects. It can be applied on a detailed 3-D road model reconstruction. The objective of this research is to develop automatic process for pole-like objects extraction from mobile lidar data. The major work includes area of interest (AOI) selection, segmentation, initial detection, features recognition and RANdom SAmple Consensus (RANSAC) pole detection. The test data is acquired by Riegl VMX-250 mobile lidar system. The test area is located at Chiu-Chung Road in Taipei city, Taiwan. The length of the test area is about 220 meters. The correctness of objects detection is about 70%. The mean error of x, y, and z coordinates objects are 0.012, 0.009, and 0.039 meters respectively. The mean error of radius is 0.011 meters. With the position and direction of objects, the extracted results can be placed in a LOD3 road model. The experimental results indicate that the proposed process may extract pole-like roadside objects effectively.

Original languageEnglish
Title of host publication34th Asian Conference on Remote Sensing 2013, ACRS 2013
PublisherAsian Association on Remote Sensing
Pages4584-4592
Number of pages9
ISBN (Print)9781629939100
StatePublished - 1 Jan 2013
Event34th Asian Conference on Remote Sensing 2013, ACRS 2013 - Bali, Indonesia
Duration: 20 Oct 201324 Oct 2013

Publication series

Name34th Asian Conference on Remote Sensing 2013, ACRS 2013
Volume5

Conference

Conference34th Asian Conference on Remote Sensing 2013, ACRS 2013
CountryIndonesia
CityBali
Period20/10/1324/10/13

Keywords

  • Mobile lidar
  • Point clouds
  • Pole-like objects extraction
  • Segmentation

Fingerprint Dive into the research topics of 'Pole-like roadside objects extraction from mobile lidar point clouds'. Together they form a unique fingerprint.

Cite this