Ultrasonic Time of Flight Diffraction Defect Recognition Based on Edge Detection

We investigate a kind of method of ultrasonic time of flight diffraction defect recognition based on edge detection in this paper. Firstly, the ultrasonic B-scan image is accumulated by successive A-scan ultrasonic signals of defect on the specimen. Secondly, we choose one A-scan ultrasonic signal as the reference signal and calculate the cross correlation function between the reference signal and other A-scan ultrasonic signals. The B-scan image is calibrated by using the largest peak of each cross correlation function. Lastly, the B-scan image is denoised by Wiener filter, and edge detection operators are applied to the Wiener-denoised B-scan image and feature of defect are detected by using these operators. Thus defect can be recognized accurately from these feature detection results of B-scan ultrasonic image. Experimental results validate this method.