relation: https://khub.utp.edu.my/scholars/9687/
title: Color-based oil spill image segmentation
creator: Osman, A.B.
creator: Ovinis, M.
creator: Ismael, M.A.
creator: Fakhruldin, M.H.
creator: Faye, I.
description: Oil spill image segmentation is an important task for quantifying the total amount of spilled oil. However, segmentation of oil region from images is a challenging problem. This is due to un-defined edge between oil and sea water. This paper proposes a novel approach based on oil color. The proposed color-based method segments the oil spill region based on clustering the colors of oil and water within the image into various groups, from which the oiled region can be segmented and defined. A k-mean cluster was applied to differentiate between the various colors. As a result, the color-based method able to determine the oiled region. A comparison to classical Otsu thresholding method showed that the color-based method had a better oil segmentation. © 2018 Author(s).
publisher: American Institute of Physics Inc.
date: 2018
type: Conference or Workshop Item
type: PeerReviewed
identifier:   Osman, A.B. and Ovinis, M. and Ismael, M.A. and Fakhruldin, M.H. and Faye, I.  (2018) Color-based oil spill image segmentation.  In: UNSPECIFIED.     
relation: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85057345191&doi=10.1063%2f1.5075603&partnerID=40&md5=f01a91553b34b759009f69a801844474
relation: 10.1063/1.5075603
identifier: 10.1063/1.5075603