Extraction of Characteristic Lines from Visualized Images of Surface Transient Displacements for the Analysis of Ultrasonic Propagation

In this paper, a feature extraction method has been presented for the non-destructive detection of defects. This method is performed through the image analysis of visualized images, obtained from 2-D scanning measurements of surface transient displacements, of ultrasonic wave propagation. The displacements induced by the ultrasonic wave on a 2-D surface have been measured and visualized by scanning surfaces of opaque media using a laser system. The ultrasonic propagation was analyzed by extracting the characteristic lines by processing the visualized images. The scattering, diffraction and mode conversion of ultrasonic waves around artificial defects of various shapes are observed using the visualization method. The visualization system provides various kinds of visualized images, such as the propagation image, amplitude image, arrival time image, velocity image, and a series of successive images as an animation of the wave propagation. The proposed method is well suited to non-destructive inspection.