Blood vessel extraction from the fundus region in the retina image is of great importance as many of the eye diseases and pathological differences can be diagnosed with this method. In this paper an automatic blood vessel extraction process is implemented with deep learning methods. The network is tested and the results are validated with standard publicly available datasets like DRIVE and compared against state or art methods, and it was observed from the analysis that the method outperforms with an average accuracy of 98.6 % which is far improved accuracy than several existing methods.
Volume 11 | 05-Special Issue
Pages: 2374-2381