Table 5. Comparison of performance evaluation results for DIARETDB1 dataset.
| Reference | Method | SE (%) |
SP (%) |
ACC (%) |
|---|---|---|---|---|
| (Das, Dandapat & Bora, 2019) | Contrast sensitivity index (CSI), Shannon entropy, multi-resolution (MR) inter-band eigen features and intra-band energy | – | – | 85.22 |
| (Long et al., 2019) | Fuzzy C-means clustering (FCM) & support vector machine (SVM) | 97.5 | 97.8 | 97.7 |
| (Mateen et al., 2020) | Feature fusion from Inception-v3, ResNet-50, and VGGNet-19 models | – | – | 98.91 |
| (Pruthi, Khanna & Arora, 2020) | Glowworm Swarm optimization | – | – | 96.56 |
| (Sharif et al., 2020) | Histogram orientation gradient (HOG) and local binary pattern (LBP) feature fusion & decision tree (DT) | 98.1 | 91.8 | 96.6 |
| (Zago et al., 2020) | Custom convolutional neural network (CNN) | 90 | 87 | – |
| (Chetoui & Akhloufi, 2020) | Extended Inception-Resnet-v2 network fine-tuned by cosine annealing strategy | 98.8 | 90.1 | 97.1 |
| (Alaguselvi & Murugan, 2020) | Morphological operation, matched filter, principal component analysis (PCA), edge detection by ISODATA, and convex hull transform | 99.03 | 98.37 | 98.68 |
| Proposed | A fusion of texture and ridgelet features & SMO | 98.87 | 95.24 | 97.05 |