Submit Manuscript  

Article Details


Fundus Image Segmentation and Feature Extraction for the Detection of Glaucoma: A New Approach

[ Vol. 14 , Issue. 1 ]

Author(s):

Syedia Tahseen Fatima Bokhari, Muhammad Sharif, Mussarat Yasmin* and Steven L. Fernandes   Pages 77 - 87 ( 11 )

Abstract:


Background: The second chief foundation of enduring visual deficiency around the world is Glaucoma. Timely exposure of this disease can confine the advancement in disease progression. Calculating Cup to Disk Ratio is amongst the effective ways for figuring out the presence of glaucoma.

Methods and Materials: In this paper, a Cup to Disk Ratio and Rim to Disk Ratio calculation based glaucoma detection method is proposed. The proposed technique comprises of three main phases. In the preliminary stage, the image is preprocessed and optic disk is segmented out using Fourier Transform, Active Contour and Circular Hough Transform followed by ocular cup segmentation stage. Ocular cup segmentation is carried out through morphological operations, Circular Hough Transform, Median Filter and Active Contour Segmentation algorithms. The outcomes of proposed technique are tested on 200 images taken from four different databases including comparison with the results of approaches proposed and implemented in the prospect of ocular cup and optic papilla based segmentation.

Results & Conclusion: The results achieved through the proposed technique showed that the proposed strategy acquired high exactness in the segmentation of optic papilla and ocular cup hence obtained RDR and CDR values near the clinical CDR values.

Keywords:

Cup to disk ratio, glaucoma detection, neuro-retinal rim, excavation detection, rim to disk ratio, ellipse fitting.

Affiliation:

Department of Computer Science, COMSATS Institute of Information Technology, Wah Cantt, Department of Computer Science, COMSATS Institute of Information Technology, Wah Cantt, Department of Computer Science, COMSATS Institute of Information Technology, Wah Cantt, Department of Electronics and Communication Engineering, Sahyadri College of Engineering & Management, Mangalore, Karnataka

Graphical Abstract:



Read Full-Text article