dc.contributor.author |
Kleebauer, M
|
|
dc.contributor.author |
Marz, Christopher A
|
|
dc.contributor.author |
Reudenbach, C
|
|
dc.contributor.author |
Braun, M
|
|
dc.date.accessioned |
2024-02-05T09:26:50Z |
|
dc.date.available |
2024-02-05T09:26:50Z |
|
dc.date.issued |
2023-12 |
|
dc.identifier.citation |
Kleebauer, M., Marz, C.A., Reudenbach, C. & Braun, M. 2023. Multi-resolution segmentation of solar photovoltaic systems using deep learning. <i>Remote Sensing, 15(24).</i> http://hdl.handle.net/10204/13565 |
en_ZA |
dc.identifier.issn |
2072-4292 |
|
dc.identifier.uri |
https://doi.org/10.3390/rs15245687
|
|
dc.identifier.uri |
http://hdl.handle.net/10204/13565
|
|
dc.description.abstract |
In the realm of solar photovoltaic system image segmentation, existing deep learning networks focus almost exclusively on single image sources both in terms of sensors used and image resolution. This often prevents the wide deployment of such networks. Our research introduces a novel approach to train a network on a diverse range of image data, spanning UAV, aerial, and satellite imagery at both native and aggregated resolutions of 0.1 m, 0.2 m, 0.3 m, 0.8 m, 1.6 m, and 3.2 m. Using extensive hyperparameter tuning, we first determined the best possible parameter combinations for the network based on the DeepLabV3 ResNet101 architecture. We then trained a model using the wide range of different image sources. The final network offers several advantages. It outperforms networks trained with single image sources in multiple test applications as measured by the F1-Score (95.27%) and IoU (91.04%). The network is also able to work with a variety of target imagery due to the fact that a diverse range of image data was used to train it. The model is made freely available for further applications. |
en_US |
dc.format |
Fulltext |
en_US |
dc.language.iso |
en |
en_US |
dc.relation.uri |
https://www.mdpi.com/2072-4292/15/24/5687 |
en_US |
dc.source |
Remote Sensing, 15(24) |
en_US |
dc.subject |
Deep learning |
en_US |
dc.subject |
Image segmentation |
en_US |
dc.subject |
Machine learning |
en_US |
dc.subject |
Object detection |
en_US |
dc.subject |
Photovoltaic plants |
en_US |
dc.subject |
Remote sensing |
en_US |
dc.subject |
Solar photovoltaic systems |
en_US |
dc.title |
Multi-resolution segmentation of solar photovoltaic systems using deep learning |
en_US |
dc.type |
Article |
en_US |
dc.description.pages |
21 |
en_US |
dc.description.note |
Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). |
en_US |
dc.description.cluster |
Smart Places |
en_US |
dc.description.impactarea |
Energy Supply and Demand |
en_US |
dc.identifier.apacitation |
Kleebauer, M., Marz, C. A., Reudenbach, C., & Braun, M. (2023). Multi-resolution segmentation of solar photovoltaic systems using deep learning. <i>Remote Sensing, 15(24)</i>, http://hdl.handle.net/10204/13565 |
en_ZA |
dc.identifier.chicagocitation |
Kleebauer, M, Christopher A Marz, C Reudenbach, and M Braun "Multi-resolution segmentation of solar photovoltaic systems using deep learning." <i>Remote Sensing, 15(24)</i> (2023) http://hdl.handle.net/10204/13565 |
en_ZA |
dc.identifier.vancouvercitation |
Kleebauer M, Marz CA, Reudenbach C, Braun M. Multi-resolution segmentation of solar photovoltaic systems using deep learning. Remote Sensing, 15(24). 2023; http://hdl.handle.net/10204/13565. |
en_ZA |
dc.identifier.ris |
TY - Article
AU - Kleebauer, M
AU - Marz, Christopher A
AU - Reudenbach, C
AU - Braun, M
AB - In the realm of solar photovoltaic system image segmentation, existing deep learning networks focus almost exclusively on single image sources both in terms of sensors used and image resolution. This often prevents the wide deployment of such networks. Our research introduces a novel approach to train a network on a diverse range of image data, spanning UAV, aerial, and satellite imagery at both native and aggregated resolutions of 0.1 m, 0.2 m, 0.3 m, 0.8 m, 1.6 m, and 3.2 m. Using extensive hyperparameter tuning, we first determined the best possible parameter combinations for the network based on the DeepLabV3 ResNet101 architecture. We then trained a model using the wide range of different image sources. The final network offers several advantages. It outperforms networks trained with single image sources in multiple test applications as measured by the F1-Score (95.27%) and IoU (91.04%). The network is also able to work with a variety of target imagery due to the fact that a diverse range of image data was used to train it. The model is made freely available for further applications.
DA - 2023-12
DB - ResearchSpace
DP - CSIR
J1 - Remote Sensing, 15(24)
KW - Deep learning
KW - Image segmentation
KW - Machine learning
KW - Object detection
KW - Photovoltaic plants
KW - Remote sensing
KW - Solar photovoltaic systems
LK - https://researchspace.csir.co.za
PY - 2023
SM - 2072-4292
T1 - Multi-resolution segmentation of solar photovoltaic systems using deep learning
TI - Multi-resolution segmentation of solar photovoltaic systems using deep learning
UR - http://hdl.handle.net/10204/13565
ER -
|
en_ZA |
dc.identifier.worklist |
27424 |
en_US |