Automatic foreground-background segmentation based on depth from coded aperture

Yu Peng Zhang*, Yong Tian Wang, Yang Li, Dong Dong Weng

*Corresponding author for this work

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

1 Citation (Scopus)

Abstract

Foreground-background segmentation is an important problem in computer vision, and it has many applications. We propose a technique of Automatic foreground-background segmentation based on depth from coded aperture. This method first calculates a coarse depth map using technique of coded aperture depth extraction, then estimate the general area of foreground. At last, in order to get the foreground, we use the GrabCut algorithm to segment the image. The entire progress is fully automatic, without any manual intervention. Experiments have proved its effectiveness. And we also did some simple examples for application.

Original languageEnglish
Title of host publication2012 Symposium on Photonics and Optoelectronics, SOPO 2012
DOIs
Publication statusPublished - 2012
Event2012 International Symposium on Photonics and Optoelectronics, SOPO 2012 - Shanghai, China
Duration: 21 May 201223 May 2012

Publication series

Name2012 Symposium on Photonics and Optoelectronics, SOPO 2012

Conference

Conference2012 International Symposium on Photonics and Optoelectronics, SOPO 2012
Country/TerritoryChina
CityShanghai
Period21/05/1223/05/12

Keywords

  • Computational Photography
  • Depth from Coded Aperture
  • GrabCut
  • foreground-background segmentation

Fingerprint

Dive into the research topics of 'Automatic foreground-background segmentation based on depth from coded aperture'. Together they form a unique fingerprint.

Cite this