Research on automatic defect localization for ultrasonic normal probe detection on railway wheel

Wheel detection is a significant guarantee for ensuring train-running safety. Ultrasonic phased array inspection technology is popularly applied in railway wheel and axle inspection at present. The objective of our study is to develop a data processing algorithm which makes use of image processing techniques in order to assist ultrasonic inspection operators in their difficult task. The research work described in this paper has thus been done with this objective in mind and it concerns a method that we have developed for automatic detection of defects. This method is based on region connection and region growing and it allows to position defects in the wheel under inspection by automatic analysis of the ultrasonic data.