Research Article
Plug Valve Surface Defects Identification through Nondestructive Testing and Fuzzy Deep-Learning Algorithm for Metal Porosity and Surface Evaluation
| S. No. | Evaluation standards | Fuzzy deep-learning output | Pixel classified corrosion image | Pixel classified crack image | Values | Values |
| 1 | Accuracy | 0.794 | 0.8068 | 2 | FN | 614 | 396 | 3 | FP | 4020 | 3952 | 4 | Fprate | 0.8627 | 0.9191 | 5 | Fvalue | 0.8814 | 0.8912 | 6 | Jaccard index | 0.788 | 0.8037 | 7 | MCC | 0.1818 | 0.1301 | 8 | Precision | 0.8108 | 0.8183 | 9 | Recall | 0.9656 | 0.9782 | 10 | TN | 640 | 348 | 11 | TP | 17226 | 17804 | 12 | TPrate | 0.9656 | 0.9782 |
|
|