A new and effective method for human retina optic disc segmentation with fuzzy clustering method based on active contour model
[ X ]
Tarih
2020
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Springer Heidelberg
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
In this paper, a new approach is proposed for localization and segmentation of the optic disc in human retina images. This new approach can find the boundary of the optic disc by an initial fuzzy clustering means algorithm. The proposed approach uses active contour model evolution based on a fuzzy clustering algorithm. The robustness of the proposed approach was evaluated with retinal imaging medical databases, such as DRIVE, STARE, DIARETDB1, and DRIONS. These bases contained images affected by different abnormalities, for example, diabetes, retinitis pigmentosa, and age-related macular degeneration AMD. A success detection rate with 100% accuracy was achieved for the DRIVE, DIRATEDB1, and DRIONS-DB databases, and 97.53% for the STARE database. For the optic disc segmentation, the method achieved an average accuracy and overlap in the range of 97.01-99.46% and 78.35-84.56% in these four databases. The result was compared with various methods in the literature, and it was concluded that the proposed method is more accurate than the other existing methods.
Açıklama
ABDULLAH, AHMED SULAIMAN/0000-0002-8578-1760
Anahtar Kelimeler
Retinal Image Analysis, Optic Disc, Fuzzy Clustering, Active Contour, Optic Disc Segmentation
Kaynak
Medical & Biological Engineering & Computing
WoS Q Değeri
Q2
Scopus Q Değeri
Q2
Cilt
58
Sayı
1
