Highlight removal method in polarimetric images based on Stokes parameters

Chen Sun, Fan Chao Yang

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

Abstract

Aiming at the reflected highlight in remote sensing, we propose a new method of removing reflected highlight in polarimetric images. This method is based on reflection physical model, only requiring four polarization images and reflection angle; the original intensities of target which is under water or other crystals can be calculated by Stokes parameters accurately. Comparing to conventional polarization methods, experimental results show that proposed method can remove the glitters more effectively and reserve the original characteristic of target as much as possible. This new method does not need to restrict the specific observation angle and time so that it has more versatility. This new method improves reflected highlight image quality and it can be considered very suitable in water and polarization remote sensing.

Original languageEnglish
Title of host publicationOptical Sensing and Imaging Technologies and Applications
EditorsDong Liu, Jin Lu, HaiMei Gong, Mircea Guina
PublisherSPIE
ISBN (Electronic)9781510623347
DOIs
StatePublished - 2018
Externally publishedYes
EventInternational Symposium on Optoelectronic Technology and Application 2018: Optical Sensing and Imaging Technologies and Applications 2018, OTA 2018 - Beijing, China
Duration: 22 May 201824 May 2018

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume10846
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

ConferenceInternational Symposium on Optoelectronic Technology and Application 2018: Optical Sensing and Imaging Technologies and Applications 2018, OTA 2018
Country/TerritoryChina
CityBeijing
Period22/05/1824/05/18

Keywords

  • Fresnel
  • Glitters
  • Polarization
  • Remote sensing
  • Stokes parameters

Fingerprint

Dive into the research topics of 'Highlight removal method in polarimetric images based on Stokes parameters'. Together they form a unique fingerprint.

Cite this