Optimizing Maximally Stable Extremal Regions (MSER) Parameters Using the Particle Swarm Optimization Algorithm

Abstract

The particle swarm optimization algorithm is a common method for finding solutions to problems that would otherwise require a brute force search. The maximally stable extremal region algorithm is a computer vision technique that can be used for object or region detection in images. This paper explores the use of the particle swarm optimization algorithm to find an acceptable set of parameters for the maximally stable extremal region algorithm. In addition, additions to the basic particle swarm optimization algorithm that allows finding a set of acceptable parameters from an infinite combination of parameters that a) do not violate bounds implicitly enforced by the maximally stable extremal region algorithm, and b) generate acceptable training and testing results are described. The output of the optimized maximally stable extremal region algorithm will be used in future work to segment potential regions of interest for image classification.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 20, 2019
Accession Number
AD1080875

Entities

People

  • Amy E. Bednar
  • Christopher T. Goodin
  • Jeremy E. Davis

Organizations

  • Engineer Research and Development Center

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Algorithms
  • Army
  • Army Corps Of Engineers
  • Artificial Intelligence
  • Autonomous Navigation
  • Computational Science
  • Computer Vision
  • Data Set
  • Data Sets
  • Detection
  • Digital Data
  • Ecology
  • Engineering
  • Engineers
  • Genetic Algorithms
  • Image Processing
  • Images
  • Information Systems
  • Optimization
  • Particle Swarm Optimization
  • Particles
  • Supervised Machine Learning
  • Topology
  • Training
  • Unmanned Vehicles

Fields of Study

  • Computer science

Readers

  • Computer Vision.
  • Operations Research

Technology Areas

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