Image segmentation and classification for fission track analysis for nuclear forensics using U-net model

Noam Elgad, Rami Babayew, Mark Last, Aryeh Weiss, Erez Gilad, Galit Katarivas Levy, Itzhak Halevy

Research output: Contribution to journalArticlepeer-review

Abstract

This study introduces a novel methodology for the detection and classification of fission track (FT) clusters in microscope images, employing state-of-the-art deep learning techniques for segmentation and classification (Elgad in nuclear forensics—fission track analysis—star segmentation and classification using deep learning, Ben-Gurion University, 2022). The U-Net model, a fully convolutional network, was used to carry out the segmentation of various star-like patterns in both single-class and multi-class scenarios.

Original languageEnglish
JournalJournal of Radioanalytical and Nuclear Chemistry
DOIs
StateAccepted/In press - 1 Jan 2024

Keywords

  • Computer vision
  • Fission track analysis
  • Holmeland security
  • Nuclear forensics
  • Safeguards investigations
  • U-Net

ASJC Scopus subject areas

  • Analytical Chemistry
  • Nuclear Energy and Engineering
  • Radiology Nuclear Medicine and imaging
  • Pollution
  • Spectroscopy
  • Public Health, Environmental and Occupational Health
  • Health, Toxicology and Mutagenesis

Fingerprint

Dive into the research topics of 'Image segmentation and classification for fission track analysis for nuclear forensics using U-net model'. Together they form a unique fingerprint.

Cite this