Adaptive Frequency Divers Filtering Techniques for Ultrasonic Analysis

Xing Li,Bilgutay, N.M.

DOI: https://doi.org/10.1109/isic.1990.128617

1990-01-01

Abstract:Ultrasonic nondestructive testing of large grained materials is limited by the ability of the detection process to distinguish the flaw signals from the backscattered grain noise. This noise often masks the flaw signal, leading to difficulties in its detection and identification. Similar problems exist in other applications such as radar and sonar, where clutter interferes with the detection of targets. In order to enhance the flaw visibility, the frequency diverse statistic filter has been developed. In the past, considerable success has been reported in signal-to-noise ratio (SNR) enhancement using this technique. However, the frequency diverse statistic filter is fairly sensitive to the filtering parameters which may cause the processing to be very intolerant. In the paper, two adaptive techniques for the frequency diverse statistic filtering are discussed. Ultrasonic flaw detection experiments are presented to demonstrate the performance of these techniques