An automatic blood cell segmentation method based on hyperspectral imaging technology

  • Qian Wang
  • , Qingli Li*
  • , Hongying Liu
  • , Mei Zhou
  • , Fangmin Guo
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Hyperspectral blood image has been utilized in biomedical field for a period of time. However, identifying and segmenting blood cells is still a tricky issue. Thus, this paper proposed a new method based on support vector machine (SVM) to solve this issue from hyperspectral images. Then post-processing of holes-filling and noise removing are performed on the segmented results to get completed cell. The experimental results proved the accuracy and accommodation for this new proposed method.

Original languageEnglish
Title of host publicationInternational Symposium on Photonics and Optoelectronics, SOPO 2015
EditorsZhiping Zhou
PublisherSPIE
ISBN (Electronic)9781628418699
DOIs
StatePublished - 2015
EventInternational Symposium on Photonics and Optoelectronics, SOPO 2015 - Shanghai, China
Duration: 22 Aug 201524 Aug 2015

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume9656
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

ConferenceInternational Symposium on Photonics and Optoelectronics, SOPO 2015
Country/TerritoryChina
CityShanghai
Period22/08/1524/08/15

Keywords

  • SVM
  • blood cell
  • hyperspectral imaging
  • segmentation

Fingerprint

Dive into the research topics of 'An automatic blood cell segmentation method based on hyperspectral imaging technology'. Together they form a unique fingerprint.

Cite this