Salient object detection in digital images based on superpixels and intrinsic features

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

1 Scopus citations

Abstract

Human visual system research and object detection in digital images have taken great relevance in the last years due to the different applications where it can be used, such as watermarking techniques to protect information, auto-focus on digital cameras, image compression, auto navigation, etc. This paper describes a saliency object detection method that uses intrinsic features of digital images: intensity, color, and texture. Also, Super Pixel Linear Iterative Clustering (SLIC) is employed to find the saliency object. The results have shown an accurate detection of the saliency object in digital images via comparison with well-known methods of the state-of-the-art. The results were evaluated through a subjective test using Mean Opinion Score (MOS).

Original languageEnglish
Title of host publicationProceedings of 2018 IEEE 9th International Conference on Dependable Systems, Services and Technologies, DESSERT 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages583-588
Number of pages6
ISBN (Electronic)9781538659038
DOIs
StatePublished - 9 Jul 2018
Event9th IEEE International Conference on Dependable Systems, Services and Technologies, DESSERT 2018 - Kyiv, Ukraine
Duration: 24 May 201827 May 2018

Publication series

NameProceedings of 2018 IEEE 9th International Conference on Dependable Systems, Services and Technologies, DESSERT 2018

Conference

Conference9th IEEE International Conference on Dependable Systems, Services and Technologies, DESSERT 2018
Country/TerritoryUkraine
CityKyiv
Period24/05/1827/05/18

Keywords

  • human visual system
  • intrinsic features
  • salient object detection
  • superpixels
  • visual attention

Fingerprint

Dive into the research topics of 'Salient object detection in digital images based on superpixels and intrinsic features'. Together they form a unique fingerprint.

Cite this