Spot Filtering Adaptive Thresholding (SFAT) Method for iris pigment spots segmentation approach / Mustafa Man ... [et al.]

Automation on iris pigment spots detection is an open issue that was highlighted by one of the previous researcher in 2016. The purpose is to detect the iris pigment spots in order to make an early prognosis regarding the eye cancer that was cause by the iris nevi. The nevi also known as a pigment s...

Full description

Saved in:
Bibliographic Details
Main Authors: Man, Mustafa, Ab Jabal, Mohamad Faizal, Wan Yussof, Wan Nural Jawahir, Hamid, Suhardi, Mohd Rahim, Mohd Shafry
Format: Conference or Workshop Item
Language:English
Published: 2018
Subjects:
Online Access:https://ir.uitm.edu.my/id/eprint/34903/1/34903.pdf
https://ir.uitm.edu.my/id/eprint/34903/
http://pice.uitm.edu.my/tesshi/
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Universiti Teknologi Mara
Language: English
Description
Summary:Automation on iris pigment spots detection is an open issue that was highlighted by one of the previous researcher in 2016. The purpose is to detect the iris pigment spots in order to make an early prognosis regarding the eye cancer that was cause by the iris nevi. The nevi also known as a pigment spots on the iris surface, which is one of the features on the iris surface. Hence, this paper has proposed Spot Filtering Adaptive Thresholding (SFAT) method in order to solve the highlighted issue. SFAT method has introduced a new thresholding intensity values and enhancement towards the colour detection algorithm. Then, the testing has been conducted on the Miles research digital iris images. The result achieved from the testing is 37.02% of the accuracy on the segmentation of the iris pigment spots on the iris surface. Moreover, increment 35.08% of accuracy on the iris pigment spots detection process compared with the previous method. The finding from the validation process towards SFAT method is the reliability of the method, which is was concerned on the complexity of the method implementation is low and the processing time of the method is less than 10 seconds in average compared with the previous method. In addition, the contribution from this study is in the medical imaging and image processing field of research.