Surface defect detection solutions for cylindrical objects

Machine vision cylinder defect detection

Cylinder surface defect detection

Testing requirements:

Real-time detection of cylindrical object surface defects, such as: more material less material, pit points, spiral, cracks, etc.

The detection of surface defects on cylindrical objects is more difficult than that on flat surfaces. Because the object being measured is not on the plane, there are very strict requirements on the illumination of machine vision light source and the depth of field of industrial lenses. Therefore, if you want to solve this kind of visual project, you usually face quite a challenge.

Currently, there are two main solutions for surface defect detection of cylindrical objects in the industry:

One solution is to use a linear array camera approach that rotates and shoots in conjunction with a linear light source, which is then combined into a flat image for analysis. This approach is much simpler in image processing and visual hardware configuration, but efficiency and rotation of the mechanism must be considered.

Another solution is to use two annular light sources to shine from each end of the product, and then 4-8 array cameras can be hung in the space where the cylindrical surface is so that the product can be photographed still. It can solve problems such as detection efficiency and rotation, but with this method, the lighting quality will definitely be reduced, the image processing will be a little more difficult, and since the method is illuminated from both ends of the product, it is not suitable for longer and larger products.

Testing requirements:

Real-time detection of cylindrical object surface defects, such as: more material less material, pit points, spiral, cracks, etc.

Description of scheme:

Through the mechanism: automatic material feeding/automatic feeding, CCD detection of the whole circle of material burr, material automatic collection and distinguish good and bad products.