Segmentation of Retinal Blood Vessel and Optic Disc in Retinal Fundus Images Using Image Processing Techniques

EasyChair Preprint no. 7086

Abstract

This research work focuses on the novel method of blood vessel segmentation is by utilizing the retinal fundus image. This includes the segmentation of exudates and hemorrhage in the retinal image. This is considered as the existing way to classify the vessels in the fundus image. Here, the proposed methodology is implemented with blood vessel segmentation and optic disc and cup for classification of attributes in order to diagnosis diabetic patients. The proposed hybrid of multi-resolution curvelet Transform with normalized graph cut techniques is used for segmentation and detecting the optic disk more accurately. Since the optic disc is often an essential preliminary process in the identification of another retinal image pathological and anatomical features. This approach detects the exact localization of the Optic disc and its boundary and segments the Blood Vessels. This approach detects the exact localization of the Optic disc and its boundary. The proposed method overcomes the accuracy of Optic disc detection and the low cost of implementation. The Optic disc detection success rate reaches 98.3 %. The Retinal Blood Vessel segmentation success rate reaches 98.3 %.

Keyphrases: blood vessels, Diabetic Retinopathy, optic cup, Optic Disc, Retinal image, Segmentation

@Booklet{EasyChair:7086,
  author = {Godfrin Pushparaj and Suganthi S. Devi},
  title = {Segmentation of Retinal Blood Vessel and Optic Disc in Retinal Fundus Images Using Image Processing Techniques},
  howpublished = {EasyChair Preprint no. 7086},

  year = {EasyChair, 2021}}