Request to Lighting Manufacturers
ovitThen there are the lighting manufacturers.
With the wide variety of evaluation requests we receive from FlexInspector, our work is becoming primarily "lighting selection".
Many manufacturers are competing to offer the best selection, but "really" only a few are actually useful.
Reasons for this are,
1) The concept of "increasing the contrast of defects" does not address a wide variety of unpredictable defects.
2) Even if the "relative" contrast of defects is increased, the product does not consistently produce the same image. This is especially fatal for methods that use fixed binarization levels.
(3) Even if the brightness distribution is discussed in terms of illumination alone, it will not be uniform because of the lens problem if the image is viewed through a lens. The problem lies in the image processing method that expects "uniformity," and in the FI method, "uniformity" is not necessary. There are defects that can be seen by non-uniformity.
(4) If the purpose is to "replace visual inspection," the illuminator should be able to produce an image as close to the visual image as possible. Using "color" may make certain defects more visible, but it also loses a lot of information.
And so on.
If you are a truly powerful illuminator manufacturer, you should not compete in terms of the abundance of your product lineup, but rather, you should offer a proposal that basically boils down to "this or that. Because "a single fluorescent lamp is sufficient" for many visual inspection sites.
>Many visual inspection sites say, "One fluorescent light is enough."
It is true that in the case of visual inspection, we can look for various defects by the light of a fluorescent lamp.
What is the difference between inspecting with the human eye and using a camera...?
When we inspect the same object every day, we naturally tilt it to an angle where defects are easily seen.
Does the human inspection have a higher detection capability by tilting the workpiece to an angle where defects are easier to see?
Is there any kind of lighting that says, "Anything goes!"? Is there such a thing as lighting?
>What is the difference between inspecting with the human eye and using a camera...?
>When we inspect the same workpiece every day, we naturally tilt it to an angle at which defects are easily seen.
>Does the human inspection have a higher detection capability by tilting the workpiece to an angle where defects are easier to see?
The differences are,
(1) Visual inspection has "hands".
(2) Visual inspection does not judge by a single image, but looks at changes in a series of images.
I think this is the point.
In this light, I think that discussing lighting alone is not the answer, and that we need to think in terms of a "system" that includes lighting, image inspection algorithms, and handling as a set.