PEKAT VISION Anomaly of Surface module is a good choice for the inspection of defects that you don’t expect to occur or don’t know what they will look like. As it can be trained on good (defect-free) images only (unsupervised learning), it is able to highlight everything on the surface of the concrete object that should not be there.
Textile is one of the examples that are perfect for using our Anomaly of Surface module. Since you cannot always expect specific types of defects to occur, it is good to train PEKAT VISION to know how the specific fabric looks when defect-free. It will then show you every flaw.
The great thing is that even when there’s a pattern on the textile, PEKAT VISION is able to learn the pattern and not see it as a defect.
In the presented use case we used a fabric with a pattern. We trained the software using only good images and in a few minutes, we got precise results which you can see in the video below where we captured the process of how PEKAT VISION detects various defects on textile in real-time.