Power Plant Classification from Remote Imaging with Deep Learning

Item Type Conference or Workshop Item (Paper)
Abstract Satellite remote imaging enables the detailed study of land use patterns on a global scale. We investigate the possibility to improve the information content of traditional land use classification by identifying the nature of industrial sites from medium-resolution remote sensing images. In this work, we focus on classifying different types of power plants from Sentinel-2 imaging data. Using a ResNet-50 deep learning model, we are able to achieve a mean accuracy of 90.0% in distinguishing 10 different power plant types and a background class. Furthermore, we are able to identify the cooling mechanisms utilized in thermal power plants with a mean accuracy of 87.5%. Our results enable us to qualitatively investigate the energy mix from Sentinel-2 imaging data, and prove the feasibility to classify industrial sites on a global scale from freely available satellite imagery.
Authors Mommert, Michael; Scheibenreif, Linus Mathias; Hanna, Joëlle & Borth, Damian
Research Team AIML Lab
Language English
Subjects computer science
other research area
HSG Classification contribution to scientific community
HSG Profile Area None
Date July 2021
Publisher IEEE
Place of Publication IEEE International Geoscience and Remote Sensing Symposium (IGARSS) 2021
Page Range 6391-6394
Event Title IEEE International Geoscience and Remote Sensing Symposium (IGARSS)
Event Location virtual
Publisher DOI https://doi.org/10.1109/IGARSS47720.2021.9553219
Official URL https://ieeexplore.ieee.org/document/9553219
Depositing User Prof. Dr. Michael Mommert
Date Deposited 20 Oct 2021 14:47
Last Modified 20 Jul 2022 17:46
URI: https://www.alexandria.unisg.ch/publications/264642

Download

[img] Text
power_plants.pdf - Accepted Version

Download (1MB)

Citation

Mommert, Michael; Scheibenreif, Linus Mathias; Hanna, Joëlle & Borth, Damian: Power Plant Classification from Remote Imaging with Deep Learning. 2021. - IEEE International Geoscience and Remote Sensing Symposium (IGARSS). - virtual.

Statistics

https://www.alexandria.unisg.ch/id/eprint/264642
Edit item Edit item
Feedback?