Developing Efficient Procedures for Automated Sinkhole Extraction from Lidar DEMs
Files
Download Full Text
Publication Date
June 2013
Abstract
Sinkhole detection in karst areas is usually difficult through remote sensing image interpretation. We present an efficient approach to extract mature sinkholes from lidar DEM. First, an adaptive Wiener filter (AWF) and hierarchical watershed segmentation (HWS) are applied to identify all local depression or potential sinkholes. Second, a hole-filling algorithm is applied to the potential sinkholes, and nine spatial features are extracted. Finally, the random forest classifier is used to select true sinkholes from all potential sinkholes. Our results show that this approach is efficient for detecting mature sinkholes from lidar data, and it can be used for risk assessment and hazard preparedness in karst areas.
Keywords
Sinkhole Detection, Karst Areas, Adaptive Wiener Filter (AWF), Hierarchal Watershed Segmentation (HWS)
Document Type
Article
Notes
Photogrammetric Engineering & Remote Sensing, Vol. 6 (2013-06-01).
Identifier
SFS0055975_00001
Recommended Citation
Miao, Xin; Qiu, Xiaomin; and Wu, Shuo-Sheng, "Developing Efficient Procedures for Automated Sinkhole Extraction from Lidar DEMs" (2013). KIP Articles. 1290.
https://digitalcommons.usf.edu/kip_articles/1290
