Abstract: In Lamb wave tomography, the limitation caused by data acquisition geometry generally results in incomplete projection data, therefore poor reconstruction quality. Image compounding is an effective method to improve the image quality. We apply image compounding in Lamb wave tomography. Incomplete data are acquired on an aluminum plate sample with machined defect using different transducers configurations, and images reconstructed using simultaneous iterative reconstruction technique (SIRT). Three different methods of image compounding are used, which are average compounding, rootmeansquare compounding and geometric mean compounding. The results show that the imaged flaw is visually closer to its real shape, and the reconstruction quality significantly improved after compounding.

Key words: image compounding
, tomography, Lamb wave