Sparse decomposition-based 3D ultrasound imaging and its application in pipeline defect testing using a multi-transducer composite array

Abstract

Internal ultrasonic transducer arrays have been widely applied to inspect pipeline defects. However, shortcomings such as lower defect detection ratios and lower defect imaging resolution persist. These problems prevent the accurate evaluation of pipeline defects. Therefore, a novel pipeline defect testing method and a 3D defect imaging method are proposed in this work. The defect testing method employs three acoustic incidence modes on an ultrasonic transducer array to inspect defects of various orientations and sizes. The 3D ultrasound defect imaging technique uses a spatial sector voxel model to form a quantitative visual image with the acoustic parameters extracted from the echo signal on the basis of the sparse decomposition algorithm. To obtain a defect location from its echo signal, the spatial geometry model of each transducer was established. Experiments were performed on a pipeline sample with different types of artificial cracks to verify the proposed technique. The 3D crack images provided quantitative crack information with high resolution. Moreover, the parameters of crack orientation, size and shape could be obtained from them. The experimental results also indicate the noise suppression property of the technique.