Enhancing Perimeter Protection using Φ-OTDR and CNN for Event Classification

Druh výsledku
článek ve sborníku
Popis
The paper introduces an enhanced method combining Φ-OTDR and CNNs for an accurate object classification for perimeter protection. The proposed approach achieves an accuracy of 91 % of common events in the optical fiber vicinity.
Klíčová slova
Convolutional neural networks
Distributed Acoustic Sensing
event classification
Perimeter Protection
Phase-sensitive Optical Time-Domain Reflectometry