Improving Vibro-acoustography Images Using 2D Wavelet Transform

Vibro-acoustography is an ultrasonic technique that has demonstrated a great potential for nondestructive inspection in the last years. During the vibro-acoustography experimental tests, tone-burst excitation mode is commonly used to drive the confocal transducer. However, wave reverberation can occur and it has a potential to affect VA image quality, since reverberation generates artifacts as bright and dark fringes. This paper proposes a method based on the 2D Wavelet transform to process the images from the traditional vibro-acoustography setup and attenuate the reverberation fringes. The vibro-acoustography tests were performed in a metal sample, which presented machined circular defects on its surface. Besides the 2D Wavelet transform method, the created images were also processed using a traditional image processing approach for comparison purposes. By means of a visual inspection of the resulting images, it becomes clear that the proposed method almost attenuated the reverberation fringes. Even though the proposed method can completely remove the artifacts, the benefit is not directly observed in terms of improving the characterization of the defect diameters, since the current configuration of the proposed method also attenuates the defect contours.