Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks

Giorgio Morales; Daniel Arteaga; Samuel G. Huaman; Joel Telles; Walther Palomino

doi:10.1109/INTERCON.2018.8526416

Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks

Giorgio Morales
, Daniel Arteaga
, Samuel G. Huaman
, Joel Telles
, Walther Palomino

National University of Engineering

Research output: Chapter in Book/Published conference output › Conference publication

5 Citations (Scopus)

Abstract

Detecting shadows in high-resolution satellite images is a challenging task due to the fact that shadows can easily be mistaken for low reflectance soil or water and that such images have limited spectral bands. In this work, we propose a semantic level shadow segmentation by using generative adversarial networks and created a dataset of pre-processed images for training, validation and test. In this way, we trained a generator network that produces shadow masks with condition on a satellite image patch and tries to fool a discriminator, which is trained to discern if a given mask comes from the ground truth or from the generator model. The results achieve an accuracy of 95.85% and a Kappa coefficient of 91.76%, which is superior to the compared methods.

Original language	English
Title of host publication	Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018
Publisher	IEEE
Number of pages	4
ISBN (Electronic)	9781538654903
DOIs	https://doi.org/10.1109/INTERCON.2018.8526416
Publication status	Published - 6 Nov 2018
Event	25th IEEE International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018 - Lima, Peru Duration: 8 Aug 2018 → 10 Aug 2018

Publication series

Name	Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018

Conference

Conference	25th IEEE International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018
Country/Territory	Peru
City	Lima
Period	8/08/18 → 10/08/18

Bibliographical note

Publisher Copyright:
© 2018 IEEE.

Keywords

end-to-end learning
Generative Adversarial Networks
satellite image
Shadow detection

Access to Document

10.1109/INTERCON.2018.8526416

Cite this

Morales, G., Arteaga, D., Huaman, S. G., Telles, J., & Palomino, W. (2018). Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks. In Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018 Article 8526416 (Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018). IEEE. https://doi.org/10.1109/INTERCON.2018.8526416

Morales, Giorgio ; Arteaga, Daniel ; Huaman, Samuel G. et al. / Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks. Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018. IEEE, 2018. (Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018).

@inproceedings{aed3b3219ce84cbfa249dd6129ceab65,

title = "Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks",

abstract = "Detecting shadows in high-resolution satellite images is a challenging task due to the fact that shadows can easily be mistaken for low reflectance soil or water and that such images have limited spectral bands. In this work, we propose a semantic level shadow segmentation by using generative adversarial networks and created a dataset of pre-processed images for training, validation and test. In this way, we trained a generator network that produces shadow masks with condition on a satellite image patch and tries to fool a discriminator, which is trained to discern if a given mask comes from the ground truth or from the generator model. The results achieve an accuracy of 95.85\% and a Kappa coefficient of 91.76\%, which is superior to the compared methods.",

keywords = "end-to-end learning, Generative Adversarial Networks, satellite image, Shadow detection",

author = "Giorgio Morales and Daniel Arteaga and Huaman, \{Samuel G.\} and Joel Telles and Walther Palomino",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 25th IEEE International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018 ; Conference date: 08-08-2018 Through 10-08-2018",

year = "2018",

month = nov,

day = "6",

doi = "10.1109/INTERCON.2018.8526416",

language = "English",

series = "Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018",

publisher = "IEEE",

booktitle = "Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018",

address = "United States",

}

Morales, G, Arteaga, D, Huaman, SG, Telles, J & Palomino, W 2018, Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks. in Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018., 8526416, Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018, IEEE, 25th IEEE International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018, Lima, Peru, 8/08/18. https://doi.org/10.1109/INTERCON.2018.8526416

Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks. / Morales, Giorgio; Arteaga, Daniel; Huaman, Samuel G. et al.
Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018. IEEE, 2018. 8526416 (Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018).

Research output: Chapter in Book/Published conference output › Conference publication

TY - GEN

T1 - Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks

AU - Morales, Giorgio

AU - Arteaga, Daniel

AU - Huaman, Samuel G.

AU - Telles, Joel

AU - Palomino, Walther

PY - 2018/11/6

Y1 - 2018/11/6

N2 - Detecting shadows in high-resolution satellite images is a challenging task due to the fact that shadows can easily be mistaken for low reflectance soil or water and that such images have limited spectral bands. In this work, we propose a semantic level shadow segmentation by using generative adversarial networks and created a dataset of pre-processed images for training, validation and test. In this way, we trained a generator network that produces shadow masks with condition on a satellite image patch and tries to fool a discriminator, which is trained to discern if a given mask comes from the ground truth or from the generator model. The results achieve an accuracy of 95.85% and a Kappa coefficient of 91.76%, which is superior to the compared methods.

AB - Detecting shadows in high-resolution satellite images is a challenging task due to the fact that shadows can easily be mistaken for low reflectance soil or water and that such images have limited spectral bands. In this work, we propose a semantic level shadow segmentation by using generative adversarial networks and created a dataset of pre-processed images for training, validation and test. In this way, we trained a generator network that produces shadow masks with condition on a satellite image patch and tries to fool a discriminator, which is trained to discern if a given mask comes from the ground truth or from the generator model. The results achieve an accuracy of 95.85% and a Kappa coefficient of 91.76%, which is superior to the compared methods.

KW - end-to-end learning

KW - Generative Adversarial Networks

KW - satellite image

KW - Shadow detection

UR - https://www.scopus.com/pages/publications/85058024726

UR - https://ieeexplore.ieee.org/document/8526416

U2 - 10.1109/INTERCON.2018.8526416

DO - 10.1109/INTERCON.2018.8526416

M3 - Conference publication

AN - SCOPUS:85058024726

T3 - Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018

BT - Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018

PB - IEEE

T2 - 25th IEEE International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018

Y2 - 8 August 2018 through 10 August 2018

ER -

Morales G, Arteaga D, Huaman SG, Telles J, Palomino W. Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks. In Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018. IEEE. 2018. 8526416. (Proceedings of the 2018 IEEE 25th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2018). doi: 10.1109/INTERCON.2018.8526416

Shadow Detection in High-Resolution Multispectral Satellite Imagery Using Generative Adversarial Networks

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this