Enhanced Ultrasonic Imaging in Concrete Structures with Spatial Apodization Filters

Abstract

Ultrasonic wave–based imaging is a popular nondestructive tool used to inspect RC structures. The inherent heterogeneity of concrete poses a great challenge for imaging incipient damage to concrete. This paper presents an algorithm for improving the quality of images obtained from a considerably reduced data set. A set of RC samples was prepared. The reflections of ultrasonic waves were recorded by placing a transmitter and a receiver on the surface of the concrete. The signals were processed to create an image of a section of the specimens, and the images were enhanced through a set of three-step cascading filters. The spatial apodization (SA) filter is first applied to control the spatial directivity of the ultrasonic transducers. At the second level, a contrast improvement (CI) filter is applied to further reduce ghosting. The third filter is the moving average–based reflection localization filter. Images produced by the algorithm are compared with those produced via conventional techniques in terms of clarity and contrast. It is demonstrated that the cascading filters produce a clearer image with a considerably smaller data set.