Fast immage processing on die castings

The automation of inspection tasks in the field of non-destructive testing has become an important issue in industrial production. Not only the possibility of reducing costs in the field of quality control by replacing manpower by machine intelligence is a good reason for the automation, moreover the objective and immediately available quality data allows a fast determination of systematic production problems. Thus, the aim is not only to find rejects but to enhance the quality by better and more objective production control. This paper is split into two parts. The first part presents a fast algorithm for the automatic inspection of die castings, which is able to find defects without any a-priori knowledge of the inspected object itself. The second part presents a new runtime system for image processing applications which makes use of distributed computation power and symmetric multi-processing.