Detect differences from good products.

General image sensors and image processing inspection systems detect "defects.
However, defects change as the inspection target changes, and it is impossible to predict in advance where the defects will appear.
FlexInspector changes the way of thinking and detects "defects that are different from good products". By memorizing the good, FlexInspector detects the defects that are different from the good.
It's easy to remember the good parts, even if the inspection target changes.
No matter where defects occur, they can be detected as "different from good" on the image.
It is not necessary to collect all possible defect samples in advance, as long as there are readily available good samples.
Come to think of it, in actual visual inspections, we are not looking for "defects," but rather for "areas that differ from good products," aren't we?