Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface Cracks
Landslide is one of the most frequent and destructive geohazards around the world. The accurate identification of potential landslides plays a vital role in the management of landslide risk. The use of unmanned aerial vehicle (UAV) techniques has recently gained much popularity in landslide assessme...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/11/1900 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850160981550825472 |
|---|---|
| author | Zhan Cheng Wenping Gong Michel Jaboyedoff Jun Chen Marc-Henri Derron Fumeng Zhao |
| author_facet | Zhan Cheng Wenping Gong Michel Jaboyedoff Jun Chen Marc-Henri Derron Fumeng Zhao |
| author_sort | Zhan Cheng |
| collection | DOAJ |
| description | Landslide is one of the most frequent and destructive geohazards around the world. The accurate identification of potential landslides plays a vital role in the management of landslide risk. The use of unmanned aerial vehicle (UAV) techniques has recently gained much popularity in landslide assessment; however, most of the current UAV-image-based landslide identifications rely upon visual inspections. In this paper, an image-analysis-based landslide identification framework is developed to detect the landslides in UAV images by recognizing the landslide boundaries and ground surface cracks. In this framework, object-oriented image analysis is undertaken to identify the potential landslide boundaries in the input UAV images and the ground surface cracks in the UAV images are recognized by an automatic ground surface crack recognition model, which is trained through a deep transfer learning strategy. With the aid of this transfer learning strategy, the crack recognition model trained can take advantage of the feature of local ground surface cracks in the concerned area and the crack recognition model that has well been developed based on the samples of ground surface cracks collected from different landslide sites. Then, the landslide boundaries and the ground surface cracks obtained are fused based on Boolean operations; the fusion results can allow for informed landslide identification in UAV Images. To illustrate the effectiveness of the proposed image-analysis-based landslide identification framework, the Heifangtai Terrace of Gansu, China, was selected as a study area, and the identification results are further validated through comparisons with the field survey results. |
| format | Article |
| id | doaj-art-bbd589cf748e4408a44ee2ed616f2027 |
| institution | OA Journals |
| issn | 2072-4292 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-bbd589cf748e4408a44ee2ed616f20272025-08-20T02:23:00ZengMDPI AGRemote Sensing2072-42922025-05-011711190010.3390/rs17111900Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface CracksZhan Cheng0Wenping Gong1Michel Jaboyedoff2Jun Chen3Marc-Henri Derron4Fumeng Zhao5Faculty of Engineering, China University of Geosciences, Wuhan 430074, ChinaFaculty of Engineering, China University of Geosciences, Wuhan 430074, ChinaInstitute of Earth Sciences, University of Lausanne, CH 1015 Lausanne, SwitzerlandSchool of Automation, China University of Geosciences, Wuhan 430074, ChinaInstitute of Earth Sciences, University of Lausanne, CH 1015 Lausanne, SwitzerlandFaculty of Engineering, China University of Geosciences, Wuhan 430074, ChinaLandslide is one of the most frequent and destructive geohazards around the world. The accurate identification of potential landslides plays a vital role in the management of landslide risk. The use of unmanned aerial vehicle (UAV) techniques has recently gained much popularity in landslide assessment; however, most of the current UAV-image-based landslide identifications rely upon visual inspections. In this paper, an image-analysis-based landslide identification framework is developed to detect the landslides in UAV images by recognizing the landslide boundaries and ground surface cracks. In this framework, object-oriented image analysis is undertaken to identify the potential landslide boundaries in the input UAV images and the ground surface cracks in the UAV images are recognized by an automatic ground surface crack recognition model, which is trained through a deep transfer learning strategy. With the aid of this transfer learning strategy, the crack recognition model trained can take advantage of the feature of local ground surface cracks in the concerned area and the crack recognition model that has well been developed based on the samples of ground surface cracks collected from different landslide sites. Then, the landslide boundaries and the ground surface cracks obtained are fused based on Boolean operations; the fusion results can allow for informed landslide identification in UAV Images. To illustrate the effectiveness of the proposed image-analysis-based landslide identification framework, the Heifangtai Terrace of Gansu, China, was selected as a study area, and the identification results are further validated through comparisons with the field survey results.https://www.mdpi.com/2072-4292/17/11/1900landslide identificationUAV imagesimage analysisdeep transfer learninglandslide boundariesground surface cracks |
| spellingShingle | Zhan Cheng Wenping Gong Michel Jaboyedoff Jun Chen Marc-Henri Derron Fumeng Zhao Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface Cracks Remote Sensing landslide identification UAV images image analysis deep transfer learning landslide boundaries ground surface cracks |
| title | Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface Cracks |
| title_full | Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface Cracks |
| title_fullStr | Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface Cracks |
| title_full_unstemmed | Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface Cracks |
| title_short | Landslide Identification in UAV Images Through Recognition of Landslide Boundaries and Ground Surface Cracks |
| title_sort | landslide identification in uav images through recognition of landslide boundaries and ground surface cracks |
| topic | landslide identification UAV images image analysis deep transfer learning landslide boundaries ground surface cracks |
| url | https://www.mdpi.com/2072-4292/17/11/1900 |
| work_keys_str_mv | AT zhancheng landslideidentificationinuavimagesthroughrecognitionoflandslideboundariesandgroundsurfacecracks AT wenpinggong landslideidentificationinuavimagesthroughrecognitionoflandslideboundariesandgroundsurfacecracks AT micheljaboyedoff landslideidentificationinuavimagesthroughrecognitionoflandslideboundariesandgroundsurfacecracks AT junchen landslideidentificationinuavimagesthroughrecognitionoflandslideboundariesandgroundsurfacecracks AT marchenriderron landslideidentificationinuavimagesthroughrecognitionoflandslideboundariesandgroundsurfacecracks AT fumengzhao landslideidentificationinuavimagesthroughrecognitionoflandslideboundariesandgroundsurfacecracks |