Table 10 Result obtained from dataset D5 (proposed fusion-based features)
Digit class | Precision | Recall | F1 score | Test samples | |
|---|---|---|---|---|---|
0 | 100.00 | 100.00 | 100.00 | 500 | |
1 | 100.00 | 99.80 | 99.90 | 500 | |
2 | 98.23 | 99.80 | 99.01 | 500 | |
3 | 99.80 | 98.80 | 99.30 | 500 | |
4 | 99.40 | 100.00 | 99.70 | 500 | |
5 | 99.80 | 99.60 | 99.70 | 500 | |
6 | 100.00 | 99.80 | 99.90 | 500 | |
7 | 100.00 | 100.00 | 100.00 | 500 | |
8 | 100.00 | 99.80 | 99.90 | 500 | |
9 | 100.00 | 99.60 | 99.80 | 500 | |
Overall recognition accuracy (%) | 99.72 | ||||