Object-Oriented Automatic Landslide Detection from High Resolution Digital Elevation Model—Opportunities and Challenges Based on a Case Study in the Polish Carpathians
Kamila Pawłuszek-Filipiak , Andrzej Borkowski
AbstractOne of the remedies for reducing the negative effects of landslide activity is landslide mapping. Landslide detection, carried out by using historical data analysis, stereoscopic photo interpretation and/or field works, is expensive, time-consuming and requires expert knowledge and experience. Automatic approaches for landslide detection can provide benefits such as increased efficiency and reduced costs and time. Many attempts have been made to automate the process of landslide identification but the key information for this process is provided by the high-resolution Digital Elevation Model (DEM) delivered from Airborne Laser Scanning (ALS) data. Having considered this, the objective of this study is to utilise the Object-Oriented Approach (OOA) and DEM for the detection of landslides. In this study, we use the results archived from Pawluszek et al. (ISPRS Int J Geo-Inf 8:321, 2019). The challenges and opportunities of automatic approaches are discussed, based on an investigation conducted in an area heavily affected by landslides. The study area is located close to Rożnów Lake, in Poland and stands out by various land uses. The automatic detection results achieved (OA = 85% and K = 0.6) indicate that there is a huge potential in automatic approaches. However, these approaches face difficulties in landslide detection due to the smoothing of typical landslide features. This situation appears for old landslides and landslides located in areas of active agricultural treatments. Besides the fuzzy delineation of the landslide extent, landslide amalgamation in the OOA results can be observed. Thus, automatic approaches still need to be developed and improved. At the current stage of the development, automatic approaches cannot replace validation based on field reconnaissance but can support an interpreter in their work.
|Publication size in sheets||0.5|
|Book||Guzzetti Fausto, Mihalić Arbanas Snježana , Reichenbach Paola, Sassa Kyoji, Bobrowsky Peter T., Takara Kaoru (eds.): Understanding and Reducing Landslide Disaster Risk, Contribution to Landslide Disaster Risk Reduction, 2021, Springer, ISBN 978-3-030-60226-0, 505 p., DOI:10.1007/978-3-030-60227-7|
|Keywords in English||Landslide, Landslide detection, Object-oriented approach, Classification, Airborne laser scanning|
|Score||= 20.0, 15-01-2021, MonographChapterAuthor|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.