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

Shapeme histogram projection and matching for partial object recognition.

Ying Shan1, Harpreet S Sawhney, Bogdan Matei

  • 1Vision Technologies Laboratory, Sarnoff Corporation, 201 Washington Road, Princeton, NJ 08540, USA. USA. yshan@sarnoff.com

IEEE Transactions on Pattern Analysis and Machine Intelligence
|March 29, 2006
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

Improved immunologic response to COVID-19 vaccine with prolonged dosing interval in haemodialysis patients.

Scandinavian journal of immunology·2022
Same author

Domain-Specific Priors and Meta Learning for Few-Shot First-Person Action Recognition.

IEEE transactions on pattern analysis and machine intelligence·2021
Same author

MR evaluation of biliary-enteric anastomotic stricture: does contrast-enhanced T1W MRC provide additional information?

Clinics and research in hepatology and gastroenterology·2011
Same author

Detection of recurrence in glioma: a comparative prospective study between Tc-99m GHA SPECT and F-18 FDG PET/CT.

Clinical nuclear medicine·2011
Same author

Role of positron emission tomography-computed tomography in staging and early chemotherapy response evaluation in children with neuroblastoma.

Indian journal of nuclear medicine : IJNM : the official journal of the Society of Nuclear Medicine, India·2011
Same author

Evaluation of the role of [18F]FDG-PET/CT and [68Ga]DOTATOC-PET/CT in differentiating typical and atypical pulmonary carcinoids.

Cancer imaging : the official publication of the International Cancer Imaging Society·2011

This study introduces an improved shapeme histogram method for recognizing partially observed 3D objects. The new approach enhances object recognition accuracy by projecting query histograms onto model representations.

Area of Science:

  • Computer Vision
  • 3D Object Recognition
  • Computational Geometry

Background:

  • Histograms of shape signatures, or shapemes, are established for 2D/3D shape matching.
  • Previous methods effectively used shapemes for complete object recognition.

Purpose of the Study:

  • To extend shapeme histograms for recognizing partially observed query objects.
  • To develop a robust method for matching incomplete object views against a database of complete models.

Main Methods:

  • Representing each model object using a collection of shapeme histograms.
  • A two-step matching process: 1) constrained projection of query histogram onto model subspace, 2) Bayesian framework for match measure computation with implicit feature selection.
  • Solving the projection step via a sampling algorithm.

Related Experiment Videos

Main Results:

  • The proposed method demonstrates superior performance in matching partially viewed range objects compared to original shapeme histograms.
  • Achieved better results than other existing approaches on a database of 243 models.
  • The Bayesian framework and implicit feature selection enhanced discrimination capabilities.

Conclusions:

  • The extended shapeme histogram approach effectively handles partial object observations in 3D recognition.
  • The novel projection and matching strategy significantly improves recognition accuracy for incomplete data.
  • This method offers a promising advancement for 3D object recognition tasks with occluded or incomplete data.