Classification Abnormalities Of Eyes Using Bag Of Features For Medical Purpose

Samsudin, Nurul Syahira (2019) Classification Abnormalities Of Eyes Using Bag Of Features For Medical Purpose. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (24 Pages) Classification Abnormalities Of Eyes Using Bag Of Features For Medical Purpose.pdf - Submitted Version Download (345kB)

Abstract

Bag of Features (BoF) is a model used to classify the eye images whether it is normal or abnormal for this project. Abnormal eye image is specified with Corneal acrus (CA) which a deposition of lipid around iris when the level of cholesterol in body is at high level. Iridology is an alternative method to detect diseases using iris’s pattern observation. Region of interest is vital because CA is appeared 0.1 to 1mm from limbus of iris. In order to get correct ROI, this project applied Daugman’s Integro differential operator (IDO) for segmentation and Daugman rubber sheet model for normalization process. Speeded up robust features is used as features extractor and K-mean clustering is used to cluster the databases. This is an automated screening system which will not feel discomfort and painless to users and patients because it is a non-invasive method. The result of this project showed that the category of eye images used as the testing set. The entire process is develop using Matlab R2018B

Actions (login required)

View Item