Segmenting medical images using mathematical morphology and the improved watershed transform.

• Main Authors: Cahyo Crysdian, Cahyo Crysdian, Md. Sap, Mohd. Noor, Selamat, Harihodin
• Format: Article
• Language: English
• Published: Penerbit UTM Press 2001
• Subjects: QA Mathematics
• Online Access: http://eprints.utm.my/id/eprint/8765/1/CahyoCrysdian2001_SegmentingMedicalImagesUsingMathematical.PDF; http://eprints.utm.my/id/eprint/8765/

An algorithm to derive simple presentation of medical image is introduced in this paper. It sounds important since the complex scene of this type of image resists any application for holding further image analysis. The algorithm consists of four processing stage i.e. gradient image, opening operation, watershed transform, and region merging. Due to the undesired characteristic of the existing operation for supporting segmentation process, an extension of the existing gradient image, opening operation and region merging procedure is proposed. Experiment has successfully been taken to some MRI skull images in which the proposed algorithm produced a simplified shape of the image object.