Image processing today plays an important role in plant automation. Our emphasis is on identification and traceability as well as on inspection and quality control.

Image processing in plant automation

Hand writing recognition for parcel identification
Hand writing recognition for parcel identification

In plant automation, industrial image processing is often used for object identification tasks performed to ensure complete monitoring and traceability. The identification properly speaking is done via 1D and 2D codes or via optical character recognition (OCR). In the majority of cases, data exchange with a control system level is required to ensure the current validity of the data stored. Image processing systems, for this reason, are often crucial in providing information for the planning of material flows and operating procedures.

Inspection and control

Often, inspection systems are also used in quality control with a view to increasing productivity. Downtimes due to the optimisation of processes and the repair of faults can be reduced to a minimum. 

Image processing systems offer a wide range of possibilities for monitoring the state of critical modules or the development of critical processes. Deviations from given norm parameters are immediately recorded and appropriate countermeasures, taken. Our references from the steel manufacturing and shipbuilding sectors include living examples hereof:

  • Position detection of steel strips in rolling mills
  • Position control prior to the welding together of steel plates in shipbuilding
  • Weld seam detection and tracking in automated welding applications

Next-generation 3D image processing

As an industrial partner in the LYNKEUS project (sponsored by the Federal Ministry for Education and Research), ECKELMANN was involved in introducing PMD (Photonic Mixer Device) technology into practice. These new cameras record 3D information on the propagation time of light for each pixel. By means of, amongst other things, PMD cameras, Eckelmann has implemented systems for filling level and volume measurement in parcels.

  • Identification for tracking & tracing (barcode, data matrix, OCR, hand writing recognition, Braille)
  • Inline image processing for process control, e.g. via parts and position detection
  • Inspection systems for product and packaging control
  • Integration into manufacturing execution system (MES)

Dr.-Ing. Johannes Stelter