Details
Originalsprache | Englisch |
---|---|
Seiten (von - bis) | 501-521 |
Seitenumfang | 21 |
Fachzeitschrift | Geomatics |
Jahrgang | 3 |
Ausgabenummer | 4 |
Publikationsstatus | Veröffentlicht - 26 Nov. 2023 |
Abstract
Detecting changes in soil micro-relief in farmland helps to understand degradation processes like sheet erosion. Using the high-resolution technique of terrestrial laser scanning (TLS), we generated point clouds of three 2 × 3 m plots on a weekly basis from May to mid-June in 2022 on cultivated farmland in Germany. Three well-known applications for eliminating vegetation points in the generated point cloud were tested: Cloth Simulation Filter (CSF) as a filtering method, three variants of CANUPO as a machine learning method, and ArcGIS PointCNN as a deep learning method, a sub-category of machine learning using deep neural networks. We assessed the methods with hard criteria such as F1 score, balanced accuracy, height differences, and their standard deviations to the reference surface, resulting in data gaps and robustness, and with soft criteria such as time-saving capacity, accessibility, and user knowledge. All algorithms showed a low performance at the initial measurement epoch, increasing with later epochs. While most of the results demonstrate a better performance of ArcGIS PointCNN, this algorithm revealed an exceptionally low performance in plot 1, which is describable by the generalization gap. Although CANUPO variants created the highest amount of data gaps, we recommend that CANUPO include colour values in combination with CSF.
ASJC Scopus Sachgebiete
- Erdkunde und Planetologie (insg.)
- Erdkunde und Planetologie (sonstige)
- Ingenieurwesen (insg.)
- Ingenieurwesen (sonstige)
- Umweltwissenschaften (insg.)
- Umweltwissenschaften (sonstige)
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
in: Geomatics, Jahrgang 3, Nr. 4, 26.11.2023, S. 501-521.
Publikation: Beitrag in Fachzeitschrift › Artikel › Forschung › Peer-Review
}
TY - JOUR
T1 - Comparative Analysis of Algorithms to Cleanse Soil Micro-Relief Point Clouds
AU - Ott, Simone
AU - Burkhard, Benjamin
AU - Harmening, Corinna
AU - Paffenholz, Jens-André
AU - Steinhoff-Knopp, Bastian
N1 - This research received funding from the Lower Saxonian State Authority for Mining, Energy and Geology of Lower Saxony (Landesamt für Bergbau, Energie und Geologie, LBEG), as it is a spin-off of the Lower Saxonian soil erosion monitoring programme (funding code: 207-4500125027).
PY - 2023/11/26
Y1 - 2023/11/26
N2 - Detecting changes in soil micro-relief in farmland helps to understand degradation processes like sheet erosion. Using the high-resolution technique of terrestrial laser scanning (TLS), we generated point clouds of three 2 × 3 m plots on a weekly basis from May to mid-June in 2022 on cultivated farmland in Germany. Three well-known applications for eliminating vegetation points in the generated point cloud were tested: Cloth Simulation Filter (CSF) as a filtering method, three variants of CANUPO as a machine learning method, and ArcGIS PointCNN as a deep learning method, a sub-category of machine learning using deep neural networks. We assessed the methods with hard criteria such as F1 score, balanced accuracy, height differences, and their standard deviations to the reference surface, resulting in data gaps and robustness, and with soft criteria such as time-saving capacity, accessibility, and user knowledge. All algorithms showed a low performance at the initial measurement epoch, increasing with later epochs. While most of the results demonstrate a better performance of ArcGIS PointCNN, this algorithm revealed an exceptionally low performance in plot 1, which is describable by the generalization gap. Although CANUPO variants created the highest amount of data gaps, we recommend that CANUPO include colour values in combination with CSF.
AB - Detecting changes in soil micro-relief in farmland helps to understand degradation processes like sheet erosion. Using the high-resolution technique of terrestrial laser scanning (TLS), we generated point clouds of three 2 × 3 m plots on a weekly basis from May to mid-June in 2022 on cultivated farmland in Germany. Three well-known applications for eliminating vegetation points in the generated point cloud were tested: Cloth Simulation Filter (CSF) as a filtering method, three variants of CANUPO as a machine learning method, and ArcGIS PointCNN as a deep learning method, a sub-category of machine learning using deep neural networks. We assessed the methods with hard criteria such as F1 score, balanced accuracy, height differences, and their standard deviations to the reference surface, resulting in data gaps and robustness, and with soft criteria such as time-saving capacity, accessibility, and user knowledge. All algorithms showed a low performance at the initial measurement epoch, increasing with later epochs. While most of the results demonstrate a better performance of ArcGIS PointCNN, this algorithm revealed an exceptionally low performance in plot 1, which is describable by the generalization gap. Although CANUPO variants created the highest amount of data gaps, we recommend that CANUPO include colour values in combination with CSF.
KW - ArcGIS PointCNN
KW - CANUPO
KW - CSF
KW - micro-relief
KW - point cloud classification
KW - soil surface
KW - terrestrial laser scanning
KW - vegetation detection
UR - http://www.scopus.com/inward/record.url?scp=85214529076&partnerID=8YFLogxK
U2 - 10.3390/geomatics3040027
DO - 10.3390/geomatics3040027
M3 - Article
VL - 3
SP - 501
EP - 521
JO - Geomatics
JF - Geomatics
IS - 4
ER -