Optical inspection for quality control

Globalisation and worldwide competition impose new challenges on the quality and cost of production lines. The Pattern Recognition Company offers and develops machine vision and pattern recognition solutions for automatic quality control in production processes. Efficient feature extraction and machine learning techniques are used to design classification systems for a variety of image-based inspection tasks.


LED inspection

We offer a machine vision system for the automatic inspection of LED images containing the following modules:

  • detection of various types of surface cracks,
  • detection of particles (e.g. dust),
  • recognition of uneven surfaces, and
  • recognition of uneven illumination.

We further offer machine vision algorithms that can even detect unknown types of LED defects.

defect detection for images of LEDs

Welding seam inspection

We offer a machine vision system for the automatic inspection of welding seams containing the following modules:

  • detection of various types of welding defects, and
  • classification into the classes OK and not OK only based on defect-free welding seams.

We further offer machine vision algorithms that can even detect unknown types of welding seam defects.

defect detection for welding seam images