|
Title: Semiautomatic Landslide Detection Using Remote Sensing and Slope Units
Accession Number: 01659444
Record Type: Component
Record URL: Availability: Find a library where document is available Abstract: Landslides induced by earthquakes and rainfall pose severe threats to the infrastructure of highways and high-speed railways. To plan an immediate emergency response, the location and scale of these landslides should be known beforehand. Traditionally, to detect multitemporal landslides induced by earthquakes and the long-term effects, along with other factors such as subsequent rainfall, one had to carry out image classification multiple times to calculate the variance information. The accuracy of that method is affected by accumulated errors from multiclassification, and the process is very time-consuming. In this paper, a semiautomatic approach is proposed for rapid mapping of multitemporal landslides. The approach can obtain the variance information of each landslide event in one detection process. In addition, slope units are introduced to separate the extracted conjoined landslides. The area of Chenjiaba, China, which is located in the highest seismic intensity zone of the Wenchuan earthquake in Beichuan and had strong rainfall 4 months after the earthquake, was selected as a case study to demonstrate the validity of this methodology. Accuracy assessment was carried out by comparing extracted landslides with a manually prepared landslide inventory map. Correctly detected were 90.1% and 94.2% of earthquake- and rainfall-induced landslides, respectively. Results show that this approach is capable of mapping temporal landslides efficiently and quickly.
Monograph Title: Monograph Accession #: 01629809
Language: English
Authors: Pagination: pp 104-110
Publication Date: 2017
ISBN: 9780309460422
Media Type: Digital/other
Features: Figures; References; Tables
TRT Terms: Geographic Terms: Subject Areas: Geotechnology; Highways
Files: TRIS, TRB, ATRI
Created Date: Feb 2 2018 2:07PM
More Articles from this Serial Issue:
|