Model-based computed tomography for nondestructive evaluation
X-ray Computer Tomography (CT) has become a routine medical diagnostic for cross-sectional imaging of the body. Yet, even with over two decades of experience in the medical field, CT technology is not routinely used in industrial settings for nondestructive evaluation (NDE). The sparsity of industrial CT systems is due not only to large initial capital costs, but also to several problems of a physical nature and a mathematical nature. In this research, we have taken a fresh look at the industrial CT imaging problem and present several major contributions to the field. Explicit relationships of the underlying mathematics, along with new properties, are summarized. Conventional methods of image reconstruction are then developed, and their strengths and limitations are established for practical industrial CT. New methods for geometry estimation from unreconstructed projection data are developed that are effective for both experimental and simulated data. A major study of image reconstruction from incomplete measurements, a common problem plaguing industrial CT, is presented. First, previously proposed techniques are studied in the context of NDE, then a class of reconstruction methods -- the domain-iterative algorithms, some of which are new -- is developed under a common framework and shown to have an infinite number of hybrid methods. Implementation of domain-iterative methods is limited by computer requirements and by alignment accuracy between the data and models. New techniques for addressing these problems are presented. A fast, parallel technique for computing Radon transform and backprojection operations is described. Software simulations of this technique are tested, while a hardware implementation is on-going in a separate project. Future research areas in industrial CT are summarized. 208 refs., 54 figs., 8 tabs.
- Research Organization:
- Lawrence Livermore National Lab., CA (USA)
- Sponsoring Organization:
- USDOE; USDOE, Washington, DC (USA)
- DOE Contract Number:
- W-7405-ENG-48
- OSTI ID:
- 5806601
- Report Number(s):
- UCRL-LR-106884; ON: DE91013462
- Resource Relation:
- Other Information: Thesis (Ph.D.)
- Country of Publication:
- United States
- Language:
- English
Similar Records
Tomographic image reconstruction and rendering with texture-mapping hardware
Beam hardening correction using iterative total variation (ITV)-based algorithm in CBCT reconstruction
Related Subjects
37 INORGANIC
ORGANIC
PHYSICAL AND ANALYTICAL CHEMISTRY
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
PHOTON COMPUTED TOMOGRAPHY
ALGORITHMS
BESSEL FUNCTIONS
COMPUTER CODES
FOURIER TRANSFORMATION
GEOMETRY
IMAGE PROCESSING
IMAGE SCANNERS
ITERATIVE METHODS
MATHEMATICS
COMPUTERIZED TOMOGRAPHY
DIAGNOSTIC TECHNIQUES
FUNCTIONS
INTEGRAL TRANSFORMATIONS
MATHEMATICAL LOGIC
PROCESSING
TOMOGRAPHY
TRANSFORMATIONS
440101* - Radiation Instrumentation- General Detectors or Monitors & Radiometric Instruments
400102 - Chemical & Spectral Procedures
990200 - Mathematics & Computers