Detecting a predefined solar spot group with a pretrained convolutional neural network

Thiago O. Camargo, Denise Pechebovicz, Sthefanie M. Premebida, Vinicios R. Soares, Virginia Baroncini, Hugo Siqueira, Diego Oliva, Marcella Martins

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Image processing and machine learning techniques have been applied in several fields of research. Astronomy is one of these areas where relevant methods have been developed to deal mainly with identifying and classification. This work introduces a new approach to characterize and classify sunspots which appear in the solar photosphere for expressing intense magnetic fields. Theses magnetic fields present significant effects on Earth. Solar images were captured in the Intensitygram Flat format, found in the Helioseismic and Magnetic Imager (HMI), of the Solar Dynamics Observatory (SDO), which facilitate processing for identification and sunspots count. This method consists of two steps: image pre-processing and training phase using convolutional neural network (CNN) to identify and characterize spots and groups. Results showed a high potential of this processing that become a competitive classifier for the sunspots groups.

Original languageEnglish
Title of host publication2019 IEEE Colombian Conference on Applications in Computational Intelligence, ColCACI 2019 - Proceedings
EditorsAlvaro David Orjuela-Canon
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728116143
DOIs
Publication statusPublished - Jun 2019
Externally publishedYes
Event2019 IEEE Colombian Conference on Applications in Computational Intelligence, ColCACI 2019 - Barranquilla, Colombia
Duration: 5 Jun 20197 Jun 2019

Publication series

Name2019 IEEE Colombian Conference on Applications in Computational Intelligence, ColCACI 2019 - Proceedings

Conference

Conference2019 IEEE Colombian Conference on Applications in Computational Intelligence, ColCACI 2019
CountryColombia
CityBarranquilla
Period5.6.197.6.19

Keywords

  • classification
  • convolutional neural network
  • Image processing
  • sunspots

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications
  • Computer Science Applications
  • Computer Vision and Pattern Recognition
  • Information Systems

Fingerprint Dive into the research topics of 'Detecting a predefined solar spot group with a pretrained convolutional neural network'. Together they form a unique fingerprint.

  • Cite this

    Camargo, T. O., Pechebovicz, D., Premebida, S. M., Soares, V. R., Baroncini, V., Siqueira, H., Oliva, D., & Martins, M. (2019). Detecting a predefined solar spot group with a pretrained convolutional neural network. In A. D. Orjuela-Canon (Ed.), 2019 IEEE Colombian Conference on Applications in Computational Intelligence, ColCACI 2019 - Proceedings [8781990] (2019 IEEE Colombian Conference on Applications in Computational Intelligence, ColCACI 2019 - Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ColCACI.2019.8781990