For the appearance defect detection of the gasket, this application is relatively small. In everyone's feeling, the gasket is worthless. If you use visual detection, the cost will be too much. Visual measurement will be considered unless it is a kind of precision product. However, even if you need to consider visual detection, it is more troublesome, because appearance defect detection itself is more troublesome.

For defect detection of such products, different defect types may require different light sources for lighting. If the surface is scratched, the low-angle ring light source may be more suitable; for the compression injury, the coaxial parallel light source, the low-angle ring light source, etc. are more suitable; if the inner and outer diameter burr, the surface light source can be considered; if the inner and outer diameter collapse edge, the low-angle annular light source can be considered for lighting. It is estimated that it is difficult to use one light source to solve all defect types. 

 The requirements for cameras are not high, because if there are defects in such hardware products, they are usually relatively large. Therefore, conventional resolution cameras can meet the requirements, such as the 1.3-megapixel camera. Conventional CCTV lenses can be used to meet the measurement requirements. It is necessary to comprehensively consider a variety of processing algorithms, such as finding table enhancement image, finding circles, coordinate system conversion, particle analysis, texture analysis, etc.

Detect the effect 

 Good product

A flat piece

Scratching

Dirty

Uneven

Change shape