OPTIMISATION OF RADIOGRAPHIC TECHNIQUES FOR AUTOMATIC EVALUATION OF REAL TIME IMAGES IN THE STEEL INDUSTRY
The development of radiographic methods in the steel industry has been restricted by the need to use photographic film as a sensor. Three real time radiographic systems are examined: - fluorescent screen and low light level TV camera - light amplifier and high definition TV camera - X-ray sensors, linear array. The performance of the different systems is compared on standard samples with artificially produced defects of well known size and shape. This was restricted to conditions for detection of inclusion clusters in hot- and cold-stripped steel up to 3 mm thick. The first system was not suitable but the latter two gave interesting results without giving the same sharpness as radiographic film. The third system (X-ray sensors linear array) is very well suited to the detection of long defects on moving products. In this case, special algorithms were developed in order to improve sharpness image and automatic detection of defects. They show the feasibility of a real time application with architectures built round the VME bus and specialised real video time devices. A simplified system, using concepts developed in this research, will be built in order to detect inclusions of cold-stripped steel samples.
Bibliographic Reference: EUR 12471 FR (1989), 105 PP., MF, ECU 8, BLOW-UP COPY ECU 13.75 THE OFFICE FOR OFFICIAL PUBLICATIONS OF THE EUROPEAN COMMUNITIES, L-2985 LUXEMBOURG
Record Number: 1989128056200 / Last updated on: 1992-08-10
Original language: FR