A Hierarchical Swarm System for Robust Real-Time Feature Detection

Master's Thesis

Abstract

In this thesis, we discuss a simple extension to the standard particle swarm optimization algorithm, inspired by genetic algorithms that allow swarms to cope better with dynamically changing fitness evaluations for a given parameter space. We demonstrate the utility of the extension in an application system for dynamical facial feature detection and tracking, which uses the proposed “real-time evolving swarms” for a continuous dynamic search of the best locations in a two-dimensional parameter space to improve upon feature detection with static parameters. We show in several experimental evaluations that the proposed method is robust to lighting changes and does not require any calibration. Moreover, the method works in real time, is computationally tractable, and not limited to the employed static feature detector, but can be applied to any n-dimensional search space.

Further, this thesis introduces a novel hierarchical extension to the standard particle swarm optimization algorithm that allows swarms to cope better with dynamically changing fitness evaluations for a given parameter space. It present the formal framework and demonstrate the utility of the extension in an application system for dynamic face detection. Specifically, the feature detector/tracker uses the proposed “hierarchical real-time swarms” for a continuous concurrent dynamic search of the best locations in a two-dimensional parameter space and the image space to improve upon feature detection and tracking in changing environments.

Attributes

Attribute NameValues
URN
  • etd-04132006-095537

Author Christopher Dennis Middendorff
Advisor Matthias Scheutz
Contributor Patrick Flynn, Committee Member
Contributor Matthias Scheutz, Committee Chair
Contributor Greg Madey, Committee Member
Degree Level Master's Thesis
Degree Discipline Computer Science and Engineering
Degree Name MSCSE
Defense Date
  • 2006-04-12

Submission Date 2006-04-13
Country
  • United States of America

Subject
  • swarm optimization

  • computer vision

  • hierarchical swarm

  • PSO

Publisher
  • University of Notre Dame

Language
  • English

Record Visibility and Access Public
Content License
  • All rights reserved

Departments and Units

Files

Please Note: You may encounter a delay before a download begins. Large or infrequently accessed files can take several minutes to retrieve from our archival storage system.