CS_Final_Metal_surface_detection

This is a final project for CoderSchool Machine Learning bootcamp on 29/12/2021.

The project is based on the dataset GC10-DET uploaded by Via.

https://www.kaggle.com/zhangyunsheng/defects-class-and-location

GC10-DET is the surface defect dataset collected in a real industry. It contains ten types of surface defects, i.e:

• Punching (Pu): 219 images.
• Weld line (Wl): 273 images.
• Crescent gap (Cg): 226 images.
• Water spot (Ws): 289 images.
• Oil spot (Os): 204 images.
• Silk spot (Ss): 650 images.
• Inclusion (In): 216 images.
• Rolled pit (Rp): 31 images.
• Crease (Cr): 52 images.
• Waist folding (Wf): 146 images.

The collected defects are on the surface of the steel sheet. The dataset includes 2280 gray-scale images and its label for type of defect and the true bounding box.