Fuzzy c-Means (FCM) for Optic Cup and Disc Segmentation with Morphological Operation

Abstract

This paper proposed the deployment of dilation and erosion with Fuzzy c-Means (FCM) as an effective optic cup and disc segmentation. The cheapest way to monitor glaucoma disease is using digital fundus camera. These images are stored in RGB format which can be split into red, green and blue channels. Previous work has identified green channel as the most suitable due to its contrast. The extracted green channel is segmented with FCM. In another test, the set of images are preprocessed with dilation and erosion to remove the vernacular. The segmentation is evaluated based on the ground truth areas that are outlined by the ophthalmologists. The CDR measurements are calculated from the diameter ratio of the segmented cup and disc. The assessment shows that omitting the vernacular area improved the sensitivity, specificity and accuracy of the segmented result.

Ähnliche Arbeiten