Table 9.
The Accuracy of 11 Types of Trash Using a Real-Time Camera with Pre-Trained ResNet18.
| Types of Trash | Number of Real Trash Tests | Total Trash is Correctly Detected | Accuracy (%) |
|---|---|---|---|
| Cardboard | 45 | 43 | 95.55555556 |
| Fabric | 17 | 15 | 88.23529412 |
| Food Packaging | 63 | 61 | 96.82539683 |
| Fruit | 21 | 20 | 95.23809524 |
| Glass | 34 | 33 | 97.05882353 |
| Leaf | 56 | 54 | 96.42857143 |
| Metal | 21 | 20 | 95.23809524 |
| Paper | 61 | 59 | 96.72131148 |
| Plastic | 52 | 50 | 96.15384615 |
| Rubber | 14 | 13 | 92.85714286 |
| Wood | 27 | 26 | 96.2962963 |
| Overall | 411 | 394 | 95.14622079 |