Developing a damage assessment model for bridge surroundings: a study of the disaster caused by Typhoon Morakot in Taiwan

Jieh Haur Chen, Mu Chun Su, Chang Yi Chen, Shih Chieh Lin

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

Typhoon Morakot has been the most severe typhoon disaster to strike Taiwan in recent decades causing tremendous damage to bridge surroundings in 2009. However, we still lack a means of assessing post-typhoon damage for follow-up rebuilding. This paper presents an integrated model that automatically measures changes in rivers, areas of damage to bridge surroundings, and changes in vegetation. The proposed model is based on a neurofuzzy mechanism enhanced by the self-organising map optimisation algorithm and also includes the particular functions of dilation, erosion, and skeletonisation to deal with river imagery. High resolution FORMOSAT-2 satellite imagery from before and after the invasion period is adopted. A bridge is randomly selected from the 129 destroyed due to the typhoon for applications of the model. The recognition results show that the river average width has increased 66% with a maximum increase of over 200%. The ruined segment of the bridge is located exactly in the most scoured region. There has also been a nearly 10% reduction in the vegetation coverage. The results yielded by the proposed model demonstrate a pinpoint accuracy rate of 99.94%. This study successfully develops a tool for large-scale damage assessment as well as for precise measurement after disasters.

Original languageEnglish
Pages (from-to)24-35
Number of pages12
JournalCivil Engineering and Environmental Systems
Volume31
Issue number1
DOIs
StatePublished - Jan 2014

Keywords

  • ANN
  • bridge
  • damage assessment
  • fuzzy
  • optimisation
  • remote sensing imagery
  • SOM
  • typhoon disaster

Fingerprint

Dive into the research topics of 'Developing a damage assessment model for bridge surroundings: a study of the disaster caused by Typhoon Morakot in Taiwan'. Together they form a unique fingerprint.

Cite this