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

A semantic approach to segmentation of overlapping objects.

T Wittenberg1, M Grobe, C Münzenmayer

  • 1Fraunhofer Institute for Integrated Circuits--Applied Electronics, Erlangen, Germany. wbg@iis.fraunhofer.de

Methods of Information in Medicine
|October 9, 2004
PubMed
Summary

This study introduces a novel semantic segmentation method for accurately separating overlapping objects in medical images. The approach effectively segments cervical cells, demonstrating its utility in complex biological and radiological analyses.

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

Comparing ensemble methods combined with different aggregating models using micrograph cell segmentation as an initial application example.

Journal of pathology informatics·2023
Same author

Number of necessary training examples for Neural Networks with different number of trainable parameters.

Journal of pathology informatics·2022
Same author

Expression of vitamin D receptors in the superior cervical ganglia of rats.

Biotechnic & histochemistry : official publication of the Biological Stain Commission·2018
Same author

Osteonecrosis of the jaw in patients transitioning from bisphosphonates to denosumab treatment for osteoporosis.

Odontology·2018
Same author

Track Q. Poster Session: Education and Training for Engineers and Physicians.

Biomedizinische Technik. Biomedical engineering·2015
Same author

Review of free software tools for image analysis of fluorescence cell micrographs.

Journal of microscopy·2014

Area of Science:

  • Computer Vision
  • Medical Image Analysis
  • Biomedical Engineering

Background:

  • Image segmentation is crucial for analyzing medical data.
  • Existing methods struggle with overlapping and occluding objects.
  • A semantic approach is needed to differentiate individual objects within complex scenes.

Purpose of the Study:

  • To introduce a novel semantic segmentation approach for overlapping objects.
  • To demonstrate its application in medical image analysis, specifically for cervical cell segmentation.
  • To provide a formal description for partitioning images into distinct semantic objects.

Main Methods:

  • Developed a novel theory enhancing full-segmentation to semantic segmentation.
  • Applied the theory to partition images into regions and then into overlapping/occluding objects.

Related Experiment Videos

  • Validated the approach on 787 PAP- and DAPI-stained micrograph image pairs of cervical cells.

    • Main Results:

    • Successfully identified, separated, and distinguished overlapping and occluding cervical cells and clusters.
    • Achieved a mean difference of 10.15% for nuclei and 10.80% for plasma segmentation compared to manual segmentation.
    • Demonstrated the ability to formally describe complex image content with overlapping objects.

    Conclusions:

    • The proposed semantic segmentation method is effective for medical image processing tasks.
    • The approach is applicable to radiology and microscopy, particularly with the subtractive transparency model.
    • Further applications can be explored by developing alternative models for handling overlaps and occlusions.