Comparison of Ultrasonic Array Imaging Algorithms for Nondestructive Evaluation

Jie Zhang*, Bruce W. Drinkwater, Paul D. Wilcox

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

69 Citations (Scopus)

Abstract

Ultrasonic array imaging algorithms have been widely used and developed in nondestructive evaluation in the last 10 years. In this paper, three imaging algorithms [total focusing method (TFM), phase-coherent imaging (PCI), and spatial compounding imaging (SCI)] are compared through both simulation and experimental measurements. In the simulation, array data sets were generated using a hybrid forward model containing a single defect among a multitude of randomly distributed point scatterers to represent backscatter from material microstructure. The number of point scatterers per unit area and their scattering amplitude were optimized to reduce computation cost. The SNR of the final images and their resolution were used to indicate the quality of the different imaging algorithms. The images of different types of defects (point reflectors and planar cracks) were used to investigate the robustness of the imaging algorithms. It is shown that PCI can yield higher image resolution and higher SNR for defects in material with weak backscatter than TFM, but that the images of cracks are distorted. Overall, TFM is the most robust algorithm across a range of different types of defects. It is also shown that the detection limit of all three imaging algorithms is almost equal for weakly scattering defects.

Original languageEnglish
Pages (from-to)1732-1745
Number of pages14
JournalIEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
Volume60
Issue number8
DOIs
Publication statusPublished - Aug 2013

Keywords

  • FULL-MATRIX
  • SCATTERING
  • REDUCTION
  • NOISE

Fingerprint

Dive into the research topics of 'Comparison of Ultrasonic Array Imaging Algorithms for Nondestructive Evaluation'. Together they form a unique fingerprint.

Cite this