Details
Originalsprache | Englisch |
---|---|
Aufsatznummer | 100001 |
Seitenumfang | 32 |
Fachzeitschrift | ISPRS Open Journal of Photogrammetry and Remote Sensing |
Jahrgang | 1 |
Frühes Online-Datum | 1 Juli 2021 |
Publikationsstatus | Veröffentlicht - Okt. 2021 |
Abstract
ASJC Scopus Sachgebiete
- Erdkunde und Planetologie (insg.)
- Computer in den Geowissenschaften
- Sozialwissenschaften (insg.)
- Geografie, Planung und Entwicklung
- Physik und Astronomie (insg.)
- Atom- und Molekularphysik sowie Optik
- Erdkunde und Planetologie (insg.)
- Erdkunde und Planetologie (sonstige)
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
in: ISPRS Open Journal of Photogrammetry and Remote Sensing, Jahrgang 1, 100001, 10.2021.
Publikation: Beitrag in Fachzeitschrift › Artikel › Forschung
}
TY - JOUR
T1 - The Hessigheim 3D (H3D) Benchmark on Semantic Segmentation of High-Resolution 3D Point Clouds and Textured Meshes from UAV LiDAR and Multi-View-Stereo
AU - Kölle, Michael
AU - Laupheimer, Dominik
AU - Schmohl, Stefan
AU - Haala, Norbert
AU - Rottensteiner, Franz
AU - Wegner, Jan Dirk
AU - Ledoux, Hugo
N1 - The H3D dataset has been captured in the context of an ongoing research project funded by the German Federal Institute of Hydrology (BfG). We would like to thank the University of Innsbruck for carrying out the flight missions. Our gratitude goes to Markus Englich for providing and maintaining H3D's IT infrastructure. We appreciate the funding of H3D as an ISPRS scientific initiative 2021 and the financial support of EuroSDR. The authors would like to show their gratitude to the State Office for Spatial Information and Land Development Baden-Wuerttemberg for providing the ALS point clouds of the village of Hessigheim. Further thanks go to Weixiao Gao from TU Delft for testing alternate segmentation approaches such as KPConv and PointNet++.
PY - 2021/10
Y1 - 2021/10
N2 - Automated semantic segmentation and object detection are of great importance in geospatial data analysis. However, supervised machine learning systems such as convolutional neural networks require large corpora of annotated training data. Especially in the geospatial domain, such datasets are quite scarce. Within this paper, we aim to alleviate this issue by introducing a new annotated 3D dataset that is unique in three ways: i) The dataset consists of both an Unmanned Aerial Vehicle (UAV) laser scanning point cloud and a 3D textured mesh. ii) The point cloud features a mean point density of about 800 pts/sqm and the oblique imagery used for 3D mesh texturing realizes a ground sampling distance of about 2-3 cm. This enables the identification of fine-grained structures and represents the state of the art in UAV-based mapping. iii) Both data modalities will be published for a total of three epochs allowing applications such as change detection. The dataset depicts the village of Hessigheim (Germany), henceforth referred to as H3D. It is designed to promote research in the field of 3D data analysis on one hand and to evaluate and rank existing and emerging approaches for semantic segmentation of both data modalities on the other hand. Ultimately, we hope that H3D will become a widely used benchmark dataset in company with the well-established ISPRS Vaihingen 3D Semantic Labeling Challenge benchmark (V3D). The dataset can be downloaded from https://ifpwww.ifp.uni-stuttgart.de/benchmark/hessigheim/default.aspx.
AB - Automated semantic segmentation and object detection are of great importance in geospatial data analysis. However, supervised machine learning systems such as convolutional neural networks require large corpora of annotated training data. Especially in the geospatial domain, such datasets are quite scarce. Within this paper, we aim to alleviate this issue by introducing a new annotated 3D dataset that is unique in three ways: i) The dataset consists of both an Unmanned Aerial Vehicle (UAV) laser scanning point cloud and a 3D textured mesh. ii) The point cloud features a mean point density of about 800 pts/sqm and the oblique imagery used for 3D mesh texturing realizes a ground sampling distance of about 2-3 cm. This enables the identification of fine-grained structures and represents the state of the art in UAV-based mapping. iii) Both data modalities will be published for a total of three epochs allowing applications such as change detection. The dataset depicts the village of Hessigheim (Germany), henceforth referred to as H3D. It is designed to promote research in the field of 3D data analysis on one hand and to evaluate and rank existing and emerging approaches for semantic segmentation of both data modalities on the other hand. Ultimately, we hope that H3D will become a widely used benchmark dataset in company with the well-established ISPRS Vaihingen 3D Semantic Labeling Challenge benchmark (V3D). The dataset can be downloaded from https://ifpwww.ifp.uni-stuttgart.de/benchmark/hessigheim/default.aspx.
KW - cs.CV
KW - UAV Laser scanning
KW - Semantic segmentation
KW - Multi-modality
KW - 3D point cloud
KW - 3D textured mesh
KW - Multi-temporality
KW - Multi-View-Stereo
UR - http://www.scopus.com/inward/record.url?scp=85182622880&partnerID=8YFLogxK
U2 - 10.1016/j.ophoto.2021.100001
DO - 10.1016/j.ophoto.2021.100001
M3 - Article
VL - 1
JO - ISPRS Open Journal of Photogrammetry and Remote Sensing
JF - ISPRS Open Journal of Photogrammetry and Remote Sensing
SN - 2667-3932
M1 - 100001
ER -