A Rule-based Track Anomaly Detection Algorithm for Maritime Force Protection

Abstract

We developed an anomaly detection tool using a Rule-based Algorithm that can detect anomalies in a set of pre-recorded tracks using their curvature, speed and weave. We devised a method that can quantify the amount of curvature in a recorded surface track. The anomaly detection tool uses the limiting values for curvature, speed and weaving provided by the user to classify a track as normal or anomalous . We tested two data sets consisting of radar tracks recorded in May and August 2007. We varied the threshold values that the tool uses. We compared the results of the tool s analysis of the data sets with a visual inspection performed by a navy combat system operator. The results of the tool s analysis were in good agreement with those of the visual inspection.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 2014
Accession Number
ADA615311

Entities

People

  • A. J. Knight
  • S. Boinepalli

Organizations

  • Defence Science and Technology Group

Tags

Communities of Interest

  • Ground and Sea Platforms
  • Human Systems
  • Materials and Manufacturing Processes
  • Sensors

DTIC Thesaurus Topics

  • Agreements
  • Algorithms
  • Anomaly Detection
  • Automatic Identification Systems
  • Change Detection
  • Classification
  • Curvature
  • Data Sets
  • Detection
  • Detectors
  • Force Protection
  • Geometry
  • Identification Systems
  • Situational Awareness
  • Training
  • Two Dimensional
  • Visual Inspection

Fields of Study

  • Computer science

Readers

  • Computer Vision.
  • Regression Analysis.
  • Sensor Fusion and Tracking Systems.