Each panel is inspected by a camera system that is installed above the material flow. Defective areas are detected where the surface pattern differs from the normal (perfect) pattern. Each defective area is classified (characterized) in light and dark defects. The operator can easily change the parameters and sorting criteria.

The GreCon system is self-learning. It can distinguish between new patterns and known patterns. For known patterns, the stored image data is simply called up from a database at the beginning of a new order; the system is immediately operational. For new patterns, two successive panels are compared. If the deviations are smaller than the preset tolerance limits, these panels are used as samples. All following panels will then be compared with the new pattern. This procedure is very easy and takes only a few minutes so that the machine can be trained to any kind of pattern in a short time period.

Defects on the top surface of each panel are determined and can be represented on the monitor if desired. The quality is determined by evaluating the data, such as defect type, size and position.