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

Local interaction fields and adaptive regularizers for surface reconstruction and image relaxation

Z Yang1, S Ma

  • 1National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing, People's Republic of China. zyang@dam.brown.edu

Network (Bristol, England)
|December 23, 1998
PubMed
Summary

Novel local interaction fields and adaptive regularizers enhance image processing by implicitly smoothing data while preserving discontinuities. This approach offers stable and noise-robust solutions for early-to-middle vision tasks.

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

[Efficacy and safety of combination immunotherapy for locally advanced esophageal carcinoma].

Zhonghua zhong liu za zhi [Chinese journal of oncology]·2026
Same author

Molecular typing, biofilm characteristics and biofilm-targeted inhibition strategies of Streptococcus equi subsp. zooepidemicus isolated from donkey endometritis.

Journal of equine veterinary science·2026
Same author

PgUGT84K2, a Panax ginseng glycosyltransferase, catalyses IBA glycosylation and enhances abiotic stress tolerance in transgenic Arabidopsis.

Plant biology (Stuttgart, Germany)·2026
Same author

Black Hole Spectroscopy and Tests of General Relativity with GW250114.

Physical review letters·2026
Same author

Bevacizumab plus chemotherapy versus chemotherapy in untreated advanced non-squamous non-small-cell lung cancer patients with brain metastases (BAP BRAIN): an open-label, randomized, multicenter, phase III study.

ESMO open·2025
Same author

[Efficacy of ruxolitinib and prognostic factors in patients with myelofibrosis stratified by age].

Zhonghua xue ye xue za zhi = Zhonghua xueyexue zazhi·2025

Area of Science:

  • Computer Vision
  • Image Processing
  • Computational Mathematics

Background:

  • Conventional image processing methods often struggle with noise and preserving fine details during smoothing.
  • Existing techniques like delta-error terms can be limited in adapting to complex local image structures.

Purpose of the Study:

  • To introduce novel local interaction fields and adaptive regularizers for improved image processing.
  • To replace the conventional delta-error term with a more adaptable local interaction field.
  • To develop methods that implicitly smooth images while preserving important discontinuities.

Main Methods:

  • Development of local interaction fields, comprising an error term and an adaptive window function.
  • Integration of adaptive regularizers designed to maintain discontinuities during smoothing.

Related Experiment Videos

  • Application of gradient descent algorithms for efficient solution finding.

    • Main Results:

    • Local interaction fields implicitly promote local flatness, leading to effective smoothing.
    • Adaptive regularizers successfully preserve discontinuities while smoothing image data.
    • The proposed methods demonstrate stability across parameter variations and robustness against noise.

    Conclusions:

    • The novel local interaction fields and adaptive regularizers provide a powerful framework for image processing and early-to-middle vision.
    • The approach offers significant advantages in handling noise and preserving image structure compared to conventional methods.
    • Gradient descent is shown to be an efficient algorithm for optimizing these new models.