Extraction and motion estimation of vehicles in single-pass airborne LiDAR data towards urban traffic analysis

Wei Yao, Stefan Hinz, Uwe Stilla

Research output: Journal article publicationJournal articleAcademic researchpeer-review

48 Citations (Scopus)

Abstract

Airborne LiDAR data are characterized by involving not only rich spatial but also temporal information. It is possible to extract vehicles with motion artifacts from single-pass airborne LiDAR data, based on which the motion state and velocity of vehicles can be identified and derived. In this paper, a complete strategy for urban traffic analysis using airborne LiDAR data is presented. An adaptive 3D segmentation method is presented to facilitate the task of vehicle extraction. The method features an ability to detect local arbitrary modes at multi scales, thereby making it particularly appropriate for partitioning complex point cloud data. Vehicle objects are then extracted by a binary classification using object-based features. Furthermore, the motion analysis for extracted vehicles is performed to distinguish between moving and stationary ones. Finally, the velocity is estimated for moving vehicles. The applicability and efficiency of the presented strategy is demonstrated and evaluated on three ALS datasets acquired for the propose of city mapping, where up to 87% of vehicles have been extracted and up to 83% of moving traffic can be recovered together with reasonable velocity estimates. It can be concluded that airborne LiDAR data can provide value-added products for traffic monitoring applications, including vehicle counts, location and velocity, along with traditional products such as building models, DEMs and vegetation models. (ISPRS).
Original languageEnglish
Pages (from-to)260-271
Number of pages12
JournalISPRS Journal of Photogrammetry and Remote Sensing
Volume66
Issue number3
DOIs
Publication statusPublished - 1 May 2011
Externally publishedYes

Keywords

  • Airborne LiDAR
  • Motion estimation
  • Three-dimensional
  • Traffic monitoring
  • Urban

ASJC Scopus subject areas

  • Atomic and Molecular Physics, and Optics
  • Engineering (miscellaneous)
  • Computer Science Applications
  • Computers in Earth Sciences

Cite this