Research Article
Aeroengine Blade Surface Defect Detection System Based on Improved Faster RCNN
Table 1
Main parameters environment.
| Items | Value |
| Hardware platform | 4-DOF motion platform | Track operating speed | 10 mm/s | Industrial camera | Daheng MER-2000-5GM/C-P | Camera resolution | 3672 × 5496 | Light source | Annular adjustable LED light source | Blade dimension | 82.6 × 36.4 × 1.6 mm | CPU | Intel core(R) i7-10700 | GPU | NVIDIA RTX3090 | OS | Windows 10 | PyTorch | 1.8.0 | CUDA | 11.1 | Batchsize | 4 | Initial learning rate | 0.001 | Weight decay | 0.0005 | Epoch | 30 |
|
|