Table 1.
Difference between DenseNet and proposed DAVS-Net.
Fig 1.
Flow diagram of the proposed method.
Fig 2.
Connectivity principle of DAVS-Net.
Fig 3.
Architecture of DAVS-Net used for vessel segmentation in our work.
Table 2.
DAVS-Net encoder-decoder I/O feature map sizes.
Where EDB, EDB-C, EDB-Cat, DDB, DDB-C, DDB-Cat represent encoder dense block, encoder dense block convolution, encoder dense block concatenation, decoder dense block, decoder dense block convolution, decoder dense block concatenation, respectively. The layer with shows that layer includes rectified linear unit (ReLU), and batch normalization (BN) after.
Table 3.
Comparison of architectural differences with similar state-of-the-art networks.
Table 4.
Summary of datasets used in the experiments.
Fig 4.
Visual results on the CHASE_DB1 dataset.
From left-to-right: input images, ground truth, result obtained by our proposed method.
Fig 5.
Visual results on the DRIVE dataset.
From left-to-right: input images, ground truth, result obtained by our proposed method.
Fig 6.
Visual results on the STARE dataset.
From left-to-right: input images, ground truth, result obtained by our proposed method.
Table 5.
Performance comparison of our proposed model on CHASE_DB1 dataset with other existing models.
Table 6.
Performance comparison of our proposed model on DRIVE dataset with other existing models.
Table 7.
Performance comparison of our proposed model on STARE database with other existing models.