ULTRASONIC FLAW DETECTION BASED ON MATHEMATICAL MORPHOLOGY

In this paper, the-deterministic and statistical properties of morphological filters and their application in ultrasonic flaw detection have studied utilizing different structuring elements. It has been shown the effectiveness of the filtering process depends on the frequency content of input signals, the combination of morphological operations, and the parameters of the structuring elements. The statistical parameters (mean, variance, and skewness) of sequential morphological operations (i.e., dilation, closing, clos-erosion, and clos-opening) are examined in order to determine the noise suppression capability of morphological filters and their biasing effects. Experimental evaluation of morphological filters for flaw detection in the presence of strong scattering echoes is presented for A- and B-scans. The processed experimental results show that morphological filters can improve flaw visibility by suppressing microstructure scattering echoes. The performance of morphological filters is compared with that of recursive median filters and ensemble averaging using experimental data. Results indicate that morphological filters perform better than recursive median filters in preserving the geometric structure of the signal and can replace ensemble averaging which requires numerous measurements.