Region and boundary segmentation of microcalcifications using seed-based region growing and mathematical morphology

• Published in: Procedia - Social and Behavioral Sciences
• Main Author: Malek A.A.; Wan Abdul Rahman W.E.Z.; Ibrahim A.; Mahmud R.; Yasiran S.S.; Jumaat A.K.
• Format: Conference paper
• Language: English
• Published: Elsevier Ltd 2010
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-78651254643&doi=10.1016%2fj.sbspro.2010.12.088&partnerID=40&md5=9e12603a3ed0a5264846813333418696

In this paper an image segmentation technique is presented by combining seed based region growing and boundary segmentation in sequential order. The first process in region growing is to identify an initial seed point. Most of region growing methods identify the seed point manually which involve human interaction. Thus, automated initial seed point identification for region growing algorithm is proposed. The boundary segmentation technique is implemented in order to improve the segmentation results. The method is tested on 50 mammogram images confirmed by a radiologist to consist microcalcifications. Experimental results show that the algorithm successfully segment the microcalcifications with accuracy of 0.94. © 2010 Elsevier Ltd.