Journal cover Journal topic
Atmospheric Measurement Techniques An interactive open-access journal of the European Geosciences Union
Journal topic

Journal metrics

Journal metrics

  • IF value: 3.400 IF 3.400
  • IF 5-year value: 3.841 IF 5-year
    3.841
  • CiteScore value: 3.71 CiteScore
    3.71
  • SNIP value: 1.472 SNIP 1.472
  • IPP value: 3.57 IPP 3.57
  • SJR value: 1.770 SJR 1.770
  • Scimago H <br class='hide-on-tablet hide-on-mobile'>index value: 70 Scimago H
    index 70
  • h5-index value: 49 h5-index 49
Discussion papers
https://doi.org/10.5194/amt-2019-206
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/amt-2019-206
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Submitted as: research article 15 Jul 2019

Submitted as: research article | 15 Jul 2019

Review status
This discussion paper is a preprint. It is a manuscript under review for the journal Atmospheric Measurement Techniques (AMT).

A Convolutional Neural Network for Classifying Cloud Particles Recorded by Imaging Probes

Georgios Touloupas1,*, Annika Lauber2,*, Jan Henneberger2, Alexander Beck2, Thomas Hofmann1, and Aurélien Lucchi1 Georgios Touloupas et al.
  • 1ETH Zurich, Institute of Machine Learning
  • 2ETH Zurich, Institute for Atmospheric and Climate Science
  • *These authors contributed equally to this work

Abstract. During typical field campaigns, millions of cloud particle images are captured with imaging probes. Our interest lies in classifying these particles in order to compute the statistics needed for understanding clouds. Given the large volume of collected data, this raises the need for an automated classification approach. Traditional classification methods that require extracting features manually (e.g. decision trees and support vector machines) show reasonable performance when trained and tested on data coming from a unique dataset. However, they often have difficulties to generalize to test sets coming from other datasets where the distribution of the features might be significantly different. In practice, we found that each new dataset requires labeling a huge amount of data by hand using those methods. Convolutional neural networks have the potential to overcome this problem due to their ability to learn complex non-linear models directly from the images instead of pre-engineered features, as well as by relying on powerful regularization techniques. We show empirically that a convolutional neural network trained on cloud particles from holographic imagers generalizes well to unseen datasets. Moreover, fine-tuning the same network with a small number (256) of training images improves the classification accuracy. Thus, the automated classification with a convolutional neural network not only reduces the hand-labeling effort for new datasets but is also no longer the main error source for the classification of small particles.

Georgios Touloupas et al.
Interactive discussion
Status: open (until 09 Sep 2019)
Status: open (until 09 Sep 2019)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
[Subscribe to comment alert] Printer-friendly Version - Printer-friendly version Supplement - Supplement
Georgios Touloupas et al.
Georgios Touloupas et al.
Viewed  
Total article views: 254 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
187 65 2 254 2 1
  • HTML: 187
  • PDF: 65
  • XML: 2
  • Total: 254
  • BibTeX: 2
  • EndNote: 1
Views and downloads (calculated since 15 Jul 2019)
Cumulative views and downloads (calculated since 15 Jul 2019)
Viewed (geographical distribution)  
Total article views: 202 (including HTML, PDF, and XML) Thereof 202 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Cited  
Saved  
No saved metrics found.
Discussed  
No discussed metrics found.
Latest update: 19 Aug 2019
Publications Copernicus
Download
Short summary
Images of cloud particles give important information to improve our understanding of microphysical cloud processes. For phase-resolved measurements, a large number of water droplets and ice crystals needs to be classified by an automated approach. In this study, a convolutional neural network was designed, which exceeds the classification ability of traditional methods and therefore shortens the analysis procedure of cloud particle images.
Images of cloud particles give important information to improve our understanding of...
Citation