Published February 28, 2022 | Version v1
Video/Audio Open

Classification of a synthetic strain rate dataset acquired with Distributed Acoustic Sensing (DAS).

Creators

  • 1. Institut Terre et Environnement de Strasbourg (ITES), CNRS UMR 7063, University of Strasbourg, FRANCE /and/ FEBUS Optics, PAU, FRANCE
  • 2. Institut Terre et Environnement de Strasbourg (ITES), CNRS UMR 7063, University of Strasbourg, FRANCE
  • 3. FEBUS Optics, PAU, FRANCE

Description

The videos show the real-time acquisition of a strain rate dataset acquired with a Distributed Acoustic Sensing (DAS). The observations are collected per bloc of 4s and are segmented in six sources: noise, pedestrian, impact, backhoe, compactor, leaks. The first video shows the source identification and segmentation using a Random Forest classifier; the second video combines a Random Markov Field to the Random Forest classifier.

Files

with_markov.mp4

Files (5.7 MB)

Name Size Download all
md5:3156fbf112c699b92efc25c6d659d055
2.8 MB Preview Download
md5:62f1d8d364bef02e8c9b484c2d365648
2.9 MB Preview Download