A Locating Algorithm of Train Wheel Flaws in Ultrasonic Images Based on Cross-Correlation and Wavelet Packet Denoising

The ultrasonic phased array inspection technology has been applied to the train wheels detection. Because of the environment and characteristics of ultrasonic phased array testing, many testing images are interfered by serious noise and clutter interference, so it is difficult to identify all the flaws and will take inspector a long time. This paper presents an automatic locating algorithm of train wheel flaws in ultrasonic images based on cross-correlation and Wavelet packet denoising. Firstly, preprocessing is acted on the images, including de-noising and adaptive threshold processing. Secondly, as the echo images of web hole have periodic feature in spatial domain, image cross-correlation theory is used for extracting the period, then the echo images of web hole is removed from the B-scan images. Finally, another threshold is needed to remove small bright areas, and then the location information of flaws is obtained. Experimental results show that the proposed algorithm can effectively locate the flaws, and the average locating accuracy is higher than 96%.