Component Supplier

Member Since 2019


Our company developed a software for industrial visual inspection and quality assurance based on AI. With its unique anomaly detection feature it can detect even unexpected and previously unseen defect. PEKAT VISION is universal - it is able to check products or input raw materials which have an unstable look or shape so it can be used with all kinds of materials such as wood, stone, metal, lacquer, castings, leather, rubber or fabrics.

Content Filed Under:



Object Detector: Antigen Tests Use Case

POSTED 01/04/2022

Until now we did not pay that much attention to our Detector module. This is a pity because it is a very useful and fast technique for defect detection.

The Detector module is trained to detect objects in the picture. The process is simple – firstly select pictures for training, then draw rectangles around the object that you want to detect and then start the training. After the training is finished, set the evaluation criteria, i.e. define under which circumstances should PEKAT VISION mark the specific picture as OK or NOK.

To show you how our Detector module works, we grabbed some bottles with reagents as we saw they were not filled perfectly.

Firstly we acquired pictures - all of them containing 4 bottles with the fluid. Then we drew rectangles around the water level – we did that with approximately 5 photos that were used for the training.

When the training was finished, we set the evaluation. As we had 4 bottles in each photo, we chose that there had to be 4 water levels in the picture for it to be marked as OK.

As stated above, our Detector is a very powerful module. It can be used for different kinds of projects – alone or in combination with other PEKAT VISION modules. We have used it among others for example for the detection of defects on plastic parts of car interior, welded parts or for completeness check.


We offer the possibility of conducting a Feasibility Study - you can contact us directly through this link
or through our website:

Find more detailed explanations of functions and video tutorials on our website
or on our youtube channel