Preprints
https://doi.org/10.5194/essd-2021-270
https://doi.org/10.5194/essd-2021-270

  19 Aug 2021

19 Aug 2021

Review status: a revised version of this preprint is currently under review for the journal ESSD.

Multi-resolution dataset for photovoltaic panel segmentation from satellite and aerial imagery

Hou Jiang1, Ling Yao1,2,3, Ning Lu1,2,3, Jun Qin1,2, Tang Liu4, Yujun Liu1,5, and Chenghu Zhou1 Hou Jiang et al.
  • 1State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, 100101, China
  • 2Southern Marine Science and Engineering Guangdong Laboratory, Guangzhou 511458, China
  • 3Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application, Nanjing Normal University, Nanjing 210023, China
  • 4School of Information Engineering, China University of Geosciences (Beijing), Beijing, 100083, China
  • 5Provincial Geomatics Center of Jiangsu, Nanjing, 210013, China

Abstract. In the context of global carbon emission reduction, solar photovoltaics (PV) is experiencing rapid development. Accurate localized PV information, including location and size, is the basis for PV regulation and potential assessment of energy sector. Automatic information extraction based on deep learning requires high-quality labelled samples that should be collected at multiple spatial resolutions and under different backgrounds due to the diversity and variable scale of PV. We established a PV dataset using satellite and aerial images with spatial resolutions of 0.8 m, 0.3 m and 0.1 m, which focus on concentrated PV, distributed ground PV and fine-grained rooftop PV, respectively. The dataset contains 3716 samples of PVs installed on shrub land, grassland, cropland, saline-alkali, and water surface, as well as flat concrete, steel tile, and brick roofs. We used this dataset to examine the model performance of different deep networks on PV segmentation, and on average an intersection over union (IoU) greater than 85 % was achieved. In addition, our experiments show that direct cross application between samples with different resolutions is not feasible, and fine-tuning of the pre-trained deep networks using target samples is necessary. The dataset can support more works on PVs for greater value, such as, developing PV detection algorithm, simulating PV conversion efficiency, and estimating regional PV potential. The dataset is available from Zenodo on the following website: https://doi.org/10.5281/zenodo.5171712 (Jiang et al. 2021).

Hou Jiang et al.

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on essd-2021-270', Anonymous Referee #1, 02 Sep 2021
    • AC1: 'Reply on RC1', Ling Yao, 04 Sep 2021
    • RC4: 'Reply on RC1', Anonymous Referee #1, 13 Sep 2021
  • RC2: 'Comment on essd-2021-270', Anonymous Referee #2, 08 Sep 2021
    • AC2: 'Reply on RC2', Ling Yao, 12 Sep 2021
  • RC3: 'Comment on essd-2021-270', Anonymous Referee #3, 08 Sep 2021
    • AC3: 'Reply on RC3', Ling Yao, 12 Sep 2021
    • RC5: 'Reply on RC3', Anonymous Referee #3, 15 Sep 2021
  • CC1: 'Comment on essd-2021-270', Mingyuan Hu, 15 Sep 2021

Hou Jiang et al.

Data sets

Multi-resolution dataset for photovoltaic panel segmentation from satellite and aerial imagery Hou Jiang, Ling Yao, Yujun Liu https://doi.org/10.5281/zenodo.5171712

Hou Jiang et al.

Viewed

Total article views: 634 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
504 108 22 634 1 4
  • HTML: 504
  • PDF: 108
  • XML: 22
  • Total: 634
  • BibTeX: 1
  • EndNote: 4
Views and downloads (calculated since 19 Aug 2021)
Cumulative views and downloads (calculated since 19 Aug 2021)

Viewed (geographical distribution)

Total article views: 561 (including HTML, PDF, and XML) Thereof 561 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 24 Oct 2021
Download
Short summary
A multi-resolution (0.8 m, 0.3 m and 0.1 m) PV dataset is established using satellite and aerial images. The dataset contains 3716 samples of PVs installed on various ground lands and rooftops. The dataset can support multi-scale PV segmentation (e.g., concentrated PV, distributed ground PV and fine-grained rooftop PV), and cross application between different resolutions (e.g., from satellite to aerial samples, vice versa), as well as other research related to PV.