| Autor: |
Aruna Vinodhini, C., Sabena, S. |
| Předmět: |
|
| Zdroj: |
Journal of Circuits, Systems & Computers; Jul2023, Vol. 32 Issue 11, p1-13, 13p |
| Abstrakt: |
Segmentation of blood vessels captured using a fundus camera is the cornerstone for the medical examination of several retinal vascular disorders. In recent research studies, vessel segmentation models focus on deep neural learning. To overlook the segmentation of the toughest retinal vessels like thin vessels, a new neural network architecture is developed based on U-Net integrated with the idea of depth-wise separable convolution and the Inception network incorporated with the sparsity of information. The developed XI-UNet network is trained and tested on DRIVE, STARE and CHASE_DB1 public datasets. The performance and the achievements of the XI-UNet network are greater compared to the prevalent methods. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
