The object recognition detects deviations in the current camera image relative to a learned structure using edge structures.  

Accordingly, the basic test sequence is as follows:

• Training the object test parameters using one or more good objects
• Testing an object and/or pattern with the features taken from the training
• Testing can be performed on stationary and moving objects
• Checking for a match with the reference provides a good/poor indication after comparison with set-value criteria
• Test results output to three control outputs:
  • OK:  Presence of the trained objects and/or pattern recognized by characteristics
  • OK: Level of agreement greater than setpoint
  • ...
  • N_OK: NO presence of trained objects with the specified characteristics
  • N_OK: Deviation from degree of conformity to setpoint
  • ...
• Output of the result information onboard via PROFINET IO, Ethernet, RS232 interface – further interfaces are available via communication modules
• "Stand-alone" mode possible using integrated DI/O
• Remote control via PROFIBUS IO, PROFINET DP (via communication module), DI/O or Ethernet
• Platform-independent, web-based interface (Internet-enabled):
  • Monitoring (live image in read mode)
  • Diagnostics (fault image, log information, ...)
  • System administration (software update, ...)
  • Error analysis for troubleshooting for faulty readings
• Triggering of external ring lights  

The overall result and the components of a measurement result can be output after a test via different interfaces, and are then available for further processing on an external system, e.g. for archiving or creating a test report.