Abstract:
Automatic vessel segmentation in fundus images helps screen, diagnose, treat, and evaluate cardiovascular and ophthalmologic diseases. Retinal vessel segmentation has been difficult due to limited well-annotated data, varying vessel sizes, and intricate vessel structures. AACA-MLA-D-UNet, a new deep learning model, uses low-level detailed information and complementary information encoded in different layers to accurately distinguish vessels from the background with low model complexity. The U-Net-based model uses the dropout dense block to preserve vessel information between convolution layers and reduce over-fitting. The adaptive atrous channel attention module in the contracting path ranks feature channels automatically. After that, the multi-level attention module integrates multi-level features extracted from the expanding path and refines features at each layer using attention mechanism. The three publicly available databases—DRIVE, STARE, and CHASE_DB1—validated the proposed method. The experimental results show that the proposed method can segment retinal vessels better with less model complexity. The proposed method can handle difficult cases and generalize well.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
