Probabilistic graphical models for flood state detection of roads combining imagery and DEM

Daniel Frey, Matthias Butenuth, Daniel Straub

Research output: Contribution to journalArticlepeer-review

17 Scopus citations

Abstract

A new system for estimating the state of roads during flooding based on probabilistic graphical models is presented. The location of the roads is given by a geographic information system, whereas the up-to-date information for the assessment of flood state is delivered by remote sensing data. Furthermore, the height information from a digital elevation model (DEM) is combined with image data to improve the accuracy of the results. The presented system is based on factor graphs, which are used to model the statistical dependence between random variables. Three different models are presented: a 1-D pixel-based model, a 2-D topology-based model considering the dependences of neighboring pixels, and a 3-D multitemporal-based model, which can deal with sequential remote sensing imagery at several points in time. The proposed models are compared to a flood simulation based only on the DEM and a maximum likelihood classification based only on the image data. A numerical evaluation demonstrates the improved performance of the three proposed models.

Original languageEnglish
Article number6179971
Pages (from-to)1051-1055
Number of pages5
JournalIEEE Geoscience and Remote Sensing Letters
Volume9
Issue number6
DOIs
StatePublished - 2012

Keywords

  • Bayesian network (BN)
  • detection
  • factor graph
  • flooding
  • probabilistic graphical model

Fingerprint

Dive into the research topics of 'Probabilistic graphical models for flood state detection of roads combining imagery and DEM'. Together they form a unique fingerprint.

Cite this