On the basis of
digital image processing and pattern recognizing techniques, a quantitative auto-
measurement system was developed to diagnose the graphitization degree of carbon steel welding
microstructure for pressure vessel and main steam pipeline after long-time running under middle temperature and middle pressure. A
graphic measurement system was programmed for the shape attribute and the granularity of
graphite by using Visual Basic.Net and
choosing appropriate algorithms. The system can transform the picture into the mode of graphic analysis by using CCD camera, and can pretreat the pictures focusing on full-linear-gray-transformation to make the picture with narrow gray-level range of the original graphite
structure clear and the graphite overhanging. The different clustering can be classified with less calculation by choosing the threshold based on the 2-D histogram. The image recognization can pick the coarse and black graphite structure by using the edge-measuring method. The experiments show that the measurement time is less than 5 minutes and the results have a good agreement with those obtained from metal-microstructure analysis.