Location

Brunswick, ME

Start Date

1-1-1992 12:00 AM

Description

While tomographic techniques for radiographic image reconstruction are well established, the development of similar methods for ultrasonic data lags far behind. One of the principal reasons for this difference, despite the general similarity of the problems, is the fact that x-rays travel in a straight path even in nonhomogeneous medium and acoustic waves will generally follow a curved ray path. This introduces an additional uncertainty (ray path) into the image reconstruction algorithm. A further complication for acoustic waves is anisotropy. In most conventional image reconstruction algorithms, it is assumed that each ray traversing a given pixel will interact with the material in exactly the same way. In anisotropic media, due to the directional dependence of material properties, this is no longer the case. In this work, we explore possible techniques for overcoming these limitations. In particular, a variational formulation, based on Fermat’s principle, is used to correct for ray bending effects. Further, with the aid of some simplifying assumptions, corrections for directional property variations due to anisotropy can be introduced into the reconstruction algorithm.

Book Title

Review of Progress in Quantitative Nondestructive Evaluation

Volume

11A

Chapter

Chapter 3: Interpretive Signal Processing and Image Reconstruction

Section

Imaging and Inversion Techniques

Pages

797-803

DOI

10.1007/978-1-4615-3344-3_102

Language

en

File Format

application/pdf

Share

COinS
 
Jan 1st, 12:00 AM

Microstructure Characterizations in Composite Media Using Ultrasonic Tomography

Brunswick, ME

While tomographic techniques for radiographic image reconstruction are well established, the development of similar methods for ultrasonic data lags far behind. One of the principal reasons for this difference, despite the general similarity of the problems, is the fact that x-rays travel in a straight path even in nonhomogeneous medium and acoustic waves will generally follow a curved ray path. This introduces an additional uncertainty (ray path) into the image reconstruction algorithm. A further complication for acoustic waves is anisotropy. In most conventional image reconstruction algorithms, it is assumed that each ray traversing a given pixel will interact with the material in exactly the same way. In anisotropic media, due to the directional dependence of material properties, this is no longer the case. In this work, we explore possible techniques for overcoming these limitations. In particular, a variational formulation, based on Fermat’s principle, is used to correct for ray bending effects. Further, with the aid of some simplifying assumptions, corrections for directional property variations due to anisotropy can be introduced into the reconstruction algorithm.