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

Multi-Target Joint Detection; Tracking and Classification Based on Marginal GLMB Filter and Belief Function Theory.

Jun Liang1,2, Minzhe Li1, Zhongliang Jing1

  • 1School of Aeronautics and Astronautics, Shanghai Jiao Tong University, Shanghai 200240, China.

Sensors (Basel, Switzerland)
|August 6, 2020
PubMed
Summary

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

Incision and plasty of the descending duodenal diverticulum using an ESD technique in a child with acute pancreatitis: a case report.

BMC pediatrics·2026
Same author

Excited-State Intramolecular Proton Transfer of Deprotonated Fisetin in Poly(vinyl alcohol) Films.

The journal of physical chemistry. B·2026
Same author

Broadly tunable continuous-wave Tm:CALYO laser operating on the <sup>3</sup>H<sub>4</sub>→<sup>3</sup>H<sub>5</sub> transition.

Optics express·2026
Same author

DIFC-Net: Diffusion-Intrinsic Feature Capture for AI-Generated Image Detection.

Sensors (Basel, Switzerland)·2026
Same author

Diode-pumped high-power 1.9 µm and 2.3 µm cascade Tm:CALGO laser.

Optics express·2026
Same author

A FEN1-EDCR dual-amplification strategy for ultrasensitive detection of CHCHD10 c.176C>T mutation.

Biosensors & bioelectronics·2026

This study introduces a novel multi-target detection, tracking, and classification method using labeled random finite sets and belief functions. The approach enhances accuracy in complex scenarios with missed detections and clutter.

Area of Science:

  • Multi-target tracking and classification
  • Probabilistic data association
  • Belief function theory

Background:

  • Accurate multi-target tracking and classification remain challenging, especially with missed detections and clutter.
  • Existing methods often struggle with joint estimation and classification under uncertainty.

Purpose of the Study:

  • To develop an integrated framework for multi-target joint detection, tracking, and classification.
  • To improve the robustness and accuracy of multi-target analysis in complex environments.
  • To leverage belief function theory for enhanced classification performance.

Main Methods:

  • A class-dependent multi-model marginal generalized labeled multi-Bernoulli (MGLMB) filter for state estimation and model probability calculation.
  • A two-level classifier based on the continuous transferable belief model (cTBM) incorporating kinematic characteristics.
Keywords:
continuous transferable belief modeljoint detectionmarginal GLMB filtermulti-modeltracking and classification

Related Experiment Videos

  • Joint estimation and classification by fusing model-dependent class beliefs from continuous state feature subspaces.

    • Main Results:

    • The proposed MGLMB filter analytically computes multi-target number, states, and model probabilities.
    • The cTBM classifier effectively utilizes kinematic information for improved classification accuracy and robustness.
    • Jointly solving estimation and classification leads to superior tracking and classification performance compared to traditional methods.

    Conclusions:

    • The proposed algorithm demonstrates significant improvements in multi-target detection, tracking, and classification over conventional techniques.
    • The integration of labeled RFS and belief function theory offers a powerful solution for complex multi-target scenarios.
    • The method shows effectiveness and superiority in simulations involving miss detection and dense clutter.