Deep-Learning Approach for McIntosh-Based Classification Of Solar Active Regions Using HMI and MDI Images

Knyazeva, Irina; Rybintsev, Andrey; Ohinko, Timur; Makarenko, N. G.

doi:10.1007/978-3-030-30425-6_28

Cited by 1 publication

(4 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Considering the full set of classes might actually lead to confusion of the CNN‐classifier between classes D , E and F , as is precisely reported in Knyazeva et al. (2020) and Palladino et al. (2022).…”

Section: Methodsmentioning

confidence: 78%

“…A Faster R-CNN model is used to detect sunspots, and Inception v3 for classification. The precision score varies between 30% and 90% amongst classes, and similarly to Knyazeva et al, 2020, confusion between classes D, E, and F of component Z appears clearly.…”

Section: Sunspot Classificationmentioning

confidence: 77%

“…First the D, E, and F classes of the Z-component were merged into a single class denoted LG for "LargeGroup," since the distinction between these classes is based on the distance between the leading sunspot and trailing sunspot of the group, and not on aspect of any sunspot in the group. Considering the full set of classes might actually lead to confusion of the CNN-classifier between classes D, E and F, as is precisely reported in Knyazeva et al (2020) and Palladino et al (2022).…”

Section: Mcintosh Classes Aggregationmentioning

confidence: 98%

“…In particular, CNNs have proven superior to previous methods for image classification (Krizhevsky et al., 2012). In sunspot classification (Knyazeva et al., 2020), used a pre‐trained deep neural network and a transfer learning approach to provide an automated McIntosh classification using MDI as well as HMI (Helioseismic and Magnetic Imager, Scherrer et al., 2012) magnetograms. They achieved an average 50% precision score, and significant confusion can be seen between the classes D , E , and F of the modified Zurich class.…”

Section: Motivation and Related Workmentioning

confidence: 99%

See 3 more Smart Citations

SunSCC: Segmenting, Grouping and Classifying Sunspots From Ground‐Based Observations Using Deep Learning

Sayez,

De Vleeschouwer,

Delouille

et al. 2023

JGR Space Physics

View full text Add to dashboard Cite

We propose a fully automated system to detect, aggregate, and classify sunspot groups according to the McIntosh scheme using ground‐based white light (WL) observations from the USET facility located at the Royal Observatory of Belgium. The sunspot detection uses a Convolutional Neural Network (CNN), trained from segmentation maps obtained with an unsupervised method based on mathematical morphology and image thresholding. Given the sunspot mask, a mean‐shift algorithm is used to aggregate individual sunspots into sunspot groups. This algorithm accounts for the area of each sunspot as well as for prior knowledge regarding the shape of sunspot group. A sunspot group, defined by its bounding box and location on the Sun, is finally fed into a CNN multitask classifier. The latter predicts the three components Z, p, and c in the McIntosh classification scheme. The tasks are organized hierarchically to mimic the dependency of the second (p) and third (c) components on the first (Z). The resulting CNN‐based segmentation is more accurate than classical unsupervised methods, with an enhancement up to 16% of F1 score in detection of the smallest sunspots, and it is robust to the presence of clouds. The automated clustering method was able to separate groups with an accuracy of 80%, when compared to hand‐made USET sunspot group catalog. The CNN‐based sunspot classifier shows comparable performances to methods using continuum as well as magnetogram images recorded by instruments on space mission. We also show that an ensemble of classifiers allows differentiating reliable and potentially incorrect predictions.

show abstract

Section: Methodsmentioning

confidence: 78%

Section: Sunspot Classificationmentioning

confidence: 77%