Boundary detection of optic disk by a modified ASM method

Abstract

A new algorithm to automatically detect the boundary of optic disk in color fundus images is proposed. The optic disk is located by principal component analysis (PCA) based model, which is employed to initialize active shape model (ASM) to detect the disk boundary. ASM is modified with two aspects: one is the self-adjusting weight in the transformation from shape space to image space; the other is exclusion of outlying points in obtaining shape parameters. The modifications make the proposed algorithm more robust and converge faster than the original ASM method, especially in the case where the edge of optic disk is weak or occluded by blood vessels. © 2003 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.