An Approximate Bayesian Extended Target Tracking Algorithm

Abstract

Extended target tracking treats clusters of detections from one or more targets as one large entity to be tracked. The simplest algorithms approximate the shape of the extended target using an ellipsoid. Modern algorithms for filtering measurements for extended target tracks are based on approximations to Bayes theorem. This memo presents a new, simple approximation to the measurement-update step of an extended target-tracking filter as well as a heuristic for track initialization. Like previous work, the filter uses a Gaussian approximation for the center of the target ellipsoid and an inverse-Wishart distribution to represent the uncertainty in the shape of the ellipsoid.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 10, 2019
Accession Number
AD1087658

Entities

People

  • David F. Crouse

Organizations

  • United States Naval Research Laboratory

Tags

Communities of Interest

  • Energy and Power Technologies
  • Weapons Technologies

DTIC Thesaurus Topics

  • Algorithms
  • Cartesian Coordinates
  • Computational Science
  • Covariance
  • Data Science
  • Distribution Functions
  • Filters
  • Gaussian Distributions
  • Information Science
  • Kalman Filters
  • Military Research
  • Normal Distribution
  • Probability Distributions
  • Random Variables
  • Simulations
  • Target Tracking
  • Two Dimensional

Fields of Study

  • Engineering

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Sensor Fusion and Tracking Systems.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms