More on how to grill an integrator:
After lighting, the next issue is resolution. A vision system can't make a decision based on just one or two pixels (partly because it's using gradients, as I described in “Technical areas to assess – lighting”.) This means it's not possible to reliably detect a defect smaller than 3 x 3 pixels. In fact you should not expect to detect anything smaller than 5 pixels in any direction.
Have your vendor tell you how many millimeters per pixel the system will provide, then compare this with your minimum defect size. For example: if you can accept a defect smaller than 0.5mm x 0.5mm, then 5 pixels should cover this distance. In other words the resolution per pixel should be 0.1mm.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment