Login
- Title:
RID – Roof Information Dataset for Computer Vision-Based Photovoltaic Potential Assessment
- Document type:
- Forschungsdaten
- Publication date:
- 11.05.2022
- Responsible:
- Krapf Sebastian
- Authors:
- Krapf Sebastian, Bogenrieder Lukas, Netzler Fabian, Balke Georg, Lienkamp Markus
- Author affiliation:
- TUM
- Publisher:
- TUM
- Identifier:
- doi:10.14459/2022mp1655470
- End date of data production:
- 06.10.2021
- Subject area:
- DAT Datenverarbeitung, Informatik
- Other subject areas:
- Remote Sensing
- Resource type:
- Abbildungen von Objekten / image of objects
- Other resource types:
- Google aerial images
- Data type:
- Bilder / images; Texte / texts
- Description:
- Labeled aerial images for semantic segmentation of roof segments and roof superstructures. Data set contains 1880 annotated buildings and 1880 roof centered images. Roof segments are labeled with azimuth value or as flat. Roof superstructure classes are pvmodule, dormer, window, ladder, chimney, shadow, tree and unknown.
- Method of data assessment:
- An initial version of labels has been annotated by university members according to a labeling guideline which is accessible through the github repository. Subsequently, masks of each image have been reviewed and adapted where necessary to improve label quality. This dataset contains reviewed masks only. Initial label quality is evaluated by an annotation experiment conducted with five labelers on 26 buildings. Masks of annotation experiment are part of this dataset. Effect of annotation errors i... »
- Links:
Code: https://github.com/TUMFTM/RID
This dataset relates to the publication: https://doi.org/10.3390/rs14102299- Key words:
- roof information, roof superstructures, roof segments, computer vision, deep learning, semantic segmentation, aerial images, remote sensing, annotation, labeling, photovoltaic potential
- Technical remarks:
- View and download (1,5 GB total, 5812 Files)
The data server also offers downloads with FTP
The data server also offers downloads with rsync (password m1655470):
rsync rsync:// m1655470@dataserv.ub.tum.de/m1655470/ - Language:
- en
- Rights:
- by-nc, http://creativecommons.org/licenses/by-nc/4.0
BibTeX