Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Wavelet packet correlation methods in biometrics.

Pablo Hennings1, Jason Thornton, Jelena Kovacević

  • 1Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, Pennsylvania 15213, USA. phh@ece.cmu.edu

Applied Optics
|March 9, 2005
PubMed
Summary

We developed wavelet packet correlation filter classifiers for improved object recognition. This method enhances biometric systems by performing classification in wavelet spaces, outperforming traditional image domain approaches.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Prospective study of vaginal samples for the detection of <i>Gardnerella vaginalis</i> and <i>G. swidsinskii/leopoldii</i> using Gardnerella agar in CO<sub>2</sub> and anaerobic atmospheres.

Microbiology spectrum·2026
Same author

Parents' Attitudes and Beliefs Towards Human Papillomavirus Vaccination.

Vaccines·2025
Same author

Effect of Continuous Positive Airway Pressure Treatment on Hearing and Inner Ear Function in Patients with Obstructive Sleep Apnoea-Original Research.

Medicina (Kaunas, Lithuania)·2025
Same author

The Influence of Health Education on Vaccination Coverage and Knowledge of the School Population Related to Vaccination and Infection Caused by the Human Papillomavirus.

Vaccines·2024
Same author

Development and Validation of an Automated Classifier to Diagnose Acute Otitis Media in Children.

JAMA pediatrics·2024
Same author

The Attitudes and Practices Regarding COVID-19 among General Practitioners from Croatia and Bosnia and Herzegovina: A Cross-Sectional Study.

Behavioral sciences (Basel, Switzerland)·2023

Area of Science:

  • Computer Vision
  • Signal Processing
  • Machine Learning

Background:

  • Traditional correlation filters are designed in the image domain, often facing limitations in optimization.
  • Image domain filter design can be suboptimal for complex pattern recognition tasks.

Purpose of the Study:

  • To introduce wavelet packet correlation filter classifiers for enhanced object recognition.
  • To improve the performance of correlation filters by designing them in wavelet spaces.

Main Methods:

  • Developed wavelet packet correlation filter classifiers.
  • Proposed a pruning algorithm using a correlation energy cost function to identify optimal wavelet spaces.
  • Implemented a match score fusion algorithm for filters across the wavelet packet tree.

Related Experiment Videos

Main Results:

  • Demonstrated improved classification performance by designing correlation filters in wavelet spaces.
  • Achieved considerable improvement over standard correlation filter algorithms in a biometric recognition system using the NIST 24 fingerprint database.

Conclusions:

  • Wavelet domain design offers superior solutions for correlation filter optimization.
  • The proposed wavelet packet correlation filter classifiers are effective for object recognition tasks, particularly in biometrics.