The invention relates to the technical field of
metal surface defect detection, and discloses a
metal surface defect detection method which comprises the following steps: detecting two circular surfaces of a circular
metal sheet, and marking specific points of a first detection surface and a second detection surface; primary
coating operation is conducted on the first detection face, a first workpiece is obtained,
rubbing is conducted, new points are marked, and a first
rubbing graph is obtained; performing secondary
coating operation on the second detection surface to obtain a second workpiece,
rubbing and marking new points to obtain a second rubbing graph; acquiring images of the two rubbing graphs, and
processing the images to form a first image and a second image; normal and abnormal areas on the detection surface are obtained according to the image, and the to-be-detected workpieces are classified after comparison; according to the number of the concave positions and the convex positions of the detection surfaces and the corresponding relation, the workpieces are classified, different defect types of the workpieces are distinguished, a corresponding
processing scheme is better selected in the production process, the production cost is reduced, and the
utilization rate of the workpieces is improved.