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

Improving performance of distribution tracking through background mismatch.

Tao Zhang1, Daniel Freedman

  • 1Computer Science Department, Rensselaer Polytechnic Institute, Troy, NY 12180, USA. zhangt3@cs.rpi.edu

IEEE Transactions on Pattern Analysis and Machine Intelligence
|February 4, 2005
PubMed
Summary
This summary is machine-generated.

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

American Epilepsy Society Clinical Practice Guideline: Infantile Epilepsy.

Epilepsy currents·2026
Same author

Evaluating Arrhythmia Risk in Patients with Chronic Obstructive Pulmonary Disease Treated with Aclidinium/Formoterol Fumarate and Other Inhaled Bronchodilators: A Post-Authorization Safety Study.

International journal of chronic obstructive pulmonary disease·2026
Same author

Implementation Science Symposium: Moving the Needle in Epilepsy Care.

Epilepsy currents·2026
Same author

Evaluating extraprostatic extension of prostate cancer: pragmatic integration of MRI and PSMA-PET/CT.

Abdominal radiology (New York)·2025
Same author

HIPI: Spatially resolved multiplexed protein expression inferred from H&E WSIs.

PLoS computational biology·2024
Same author

Quality assessment of expedited AI generated reformatted images for ED acquired CT abdomen and pelvis imaging.

Abdominal radiology (New York)·2024

This study introduces a robust density matching method for tracking nonrigid objects. By incorporating background mismatching, the new approach enhances tracking accuracy and stability, especially for objects with diffuse boundaries.

Area of Science:

  • Computer Vision
  • Image Processing
  • Computational Mathematics

Background:

  • Object tracking is crucial in computer vision.
  • Existing density-matching trackers are sensitive to initial conditions and model density.
  • Nonrigid object tracking with diffuse boundaries remains challenging.

Purpose of the Study:

  • To develop a more robust density matching method for nonrigid object tracking.
  • To overcome the limitations of existing trackers regarding sensitivity and boundary handling.

Main Methods:

  • Proposed a novel density matching approach incorporating background mismatching.
  • Extended the original density-matching tracker by adding a background mismatch term to the optimization.
  • Implemented the tracker using a partial differential equation within the level-set framework.

Related Experiment Videos

Main Results:

  • The new method significantly improves robustness by reducing sensitivity to initial placements and model density.
  • Demonstrated enhanced performance in tracking objects with smooth or diffuse boundaries.
  • Experimental results on synthetic and real image sequences validate the effectiveness and robustness compared to existing methods.

Conclusions:

  • The proposed background mismatching density matching method offers a robust solution for nonrigid object tracking.
  • The level-set implementation effectively handles complex object boundaries.
  • This approach advances the state-of-the-art in visual object tracking.