PoS - Proceedings of Science
Volume 358 - 36th International Cosmic Ray Conference (ICRC2019) - CRI - Cosmic Ray Indirect
Particle Identification in Smartphone Camera Images Using the Distributed Electronic Cosmic-ray Observatory
J. Vandenbroucke, M. Meehan, J. Peacock, J. Bourbeau, F. Campos, C. Schneider, M. Winter, A. Pizzuto*, R. Morgan, B. Gallay and A.L. Simons
Full text: pdf
Pre-published on: September 02, 2019
Published on: July 02, 2021
The Distributed Electronic Cosmic-ray Observatory (DECO) is a global network of smartphones that searches images for evidence of cosmic rays and other ionizing charged particles. DECO was released to the public in 2014 and has citizen scientists from 80 countries on all seven continents participating in the project. We previously demonstrated that tracks seen in the DECO data set are caused by cosmic-ray muons by comparing the track length distribution of candidate events in the data set to the expected distribution from a cosmic-ray flux. However, robust particle identification necessary to separate cosmic rays from the radioactive background on an event-by-event basis had not previously been identified. We present a deep learning, computer vision algorithm for identifying and classifying charged particles in camera image sensors. The convolutional neural network was trained using images from the DECO data set and achieves classification performance comparable to human quality across four distinct DECO event classifications. We apply our model to the entire DECO data set and determine a selection that achieves a purity of $95\%$ when applied to cosmic-ray muons and $\ge 90\%$ for all event types. The automated classification is run on the public DECO data set in real time in order to provide classified particle interaction images to users of the app and other interested members of the public. The model and techniques used to develop it are applicable to other smartphone-based cosmic-ray detectors and data sets consisting of images of charged particles from solid-state camera sensors.
DOI: https://doi.org/10.22323/1.358.0390
How to cite

Metadata are provided both in "article" format (very similar to INSPIRE) as this helps creating very compact bibliographies which can be beneficial to authors and readers, and in "proceeding" format which is more detailed and complete.

Open Access
Creative Commons LicenseCopyright owned by the author(s) under the term of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.