Automatic segmentation of optic discussing a fine snakes in gradient vector field

Dey, Sidhartha and Tahiliani, Kapil and Kumar, Harish J R and Pediredla, A K and Seelamantula, Chandra Sekhar (2019) Automatic segmentation of optic discussing a fine snakes in gradient vector field. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2019, 12/05/2019, Brighton Center, Brighton, United Kingdom.

[img] PDF
1392.pdf - Published Version
Restricted to Registered users only

Download (773kB) | Request a copy


The optic disc is one of the prominent features of a retinal fundus image, and its segmentation is a critical component in automated retinal screening systems for ophthalmic anomalies, such as diabetic retinopathy and glaucoma. In this paper, we propose a novel method for optic disc segmentation using affine snakes, where the snake evolves using an affine transformation and requires a priori knowledge of the desired objectshape. Wedeterminetheaffinetransformationparametersbyfirstcomputingaforcefieldontheimageandthendeforming the snake till the net force on the snake is zero. The affine snakes technique excels in its speed of convergence. This is attributed to the fact that only six parameters require optimization,thesixparametersbeingthehorizontalandverticalscaling,shearingandtranslationcomponentsofanaffine transformation. Localization of the optic disc is done using normalized cross-correlation and segmentation is done using the affine snakes technique. This technique is tested on publiclyavailablefundusimagedatasets,suchasIDRiD,DrishtiGS, RIM-ONE, DRIONS-DB, and Messidor, with Dice Indices of 0.943, 0.958, 0.933, 0.913, and 0.912, respectively

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Affine snakes, optic disc, segmentation, affine transformation, gradient vector field
Subjects: Engineering > MIT Manipal > Electrical and Electronics
Depositing User: MIT Library
Date Deposited: 03 Jul 2019 08:56
Last Modified: 03 Jul 2019 08:56

Actions (login required)

View Item View Item