الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 150 |  |  |
| | | from | 150 | | |
Abstract
Three-Dimensional reconstruction of objects is a generally scientific problem and core
technology of wide variety of fields, such as Computer Aided Geometric Design (CAGD),
computer graphics, computer animation, computer vision, medical imaging, industrial purposes,
computational science, digital media, etc.
Computerized Tomography (CT) is a non-destructive technique for material testing. The
tomography imaging is a very essential technique for medical diagnostics and industrial
investigation, as well. It gives a cross sectional images for the internal structure of the examined
probe. The quality of the obtained images depends on several parameters such as physical
phenomenal measured, detection system, scanning parameters, types of radiation, image
reconstruction techniques, …etc. For transmission computed tomography, the transmitted beam is
measured for each profile. Neutrons, gamma rays, and x-rays can be used as radiation sources. In
case of medical applications, X-ray source is the most used because of the radiation hazard.
Neutron Computerized Tomography (NCT) is an image reconstruction of 2D images of an
object non-destructively radiographed at different rotational angles. The object is radiographed by
the ray passing through it. The reconstruction algorithms of physical interaction of that ray yields
3D image visualization or tomographic image. Neutron Tomography (NT) has been applied as a
novel technique for 3D visualization and characterization of internal structures of objects. Whereas
X-ray Tomography (XT) maps regions of different densities within an object, NT is sensitive to
differences in the concentration of some light materials like hydrogen, water or oil. The image
quality of NT is strongly influenced by the beam characterization and object materials. The
capability of the novel commissioned imaging technique at Egypt Second Research Reactor
(ETRR-2) to visualize separate defects has to be promoted, so advanced imaging processing
algorithms must be applied for imaging enhancement and improvement. The net results will yield
high quality 3D image formation
Three-dimensional Volume rendering (3DVR) faces many obstacles that directly affect the
quality of the reconstructed images and the rendered volume. Among these problems that occur in
the early stages due to the image acquisition system, and those occur after the stages immediately
preceding the reconstruction process. Failure to address these problems results in poor-qualitydefective rendered volume, which makes it difficult to extract sufficient information about the
scanned object.
The proposed method in this thesis presents a new technique that depends mainly on studying
and treating the causes of blurring in the reconstructed images in general and in particular the
reconstructed images resulting from the neutron tomography unit of the Egypt Second Research
Reactor (ETRR-2). The previous methods, improving the quality of the reconstruction process has
been dealt with using some filters in all stages of the NCT to raise the efficiency and show some
details regardless of the real reason behind this low quality of the reconstruction process.
Therefore, in the proposed method, the focus is on dealing with normalized images as the basis for
obtaining high and continuous quality in all stages of computerized tomography to obtain finally
a three-dimensional volume rendering with more clear details, which is almost equal to the scanned
object. Also, in the reconstruction process, the mathematical reason behind the appearance of the
reconstructed images was stained, shaky, featureless vision and had a black view. The proposed
reconstructed images were completely improved and showed high quality with clear details
compared to the previous methods used in many tomography facilities around the world.
In this work different natural samples from culture heritage, industrial samples, zoological
and botanical were scanned with different exposure times, different rotation scans with different
equiangular steps and different reactor powers. All these different parameters used to prove the
efficiency for the proposed 3DVR method without wasting time in improving its stages by filtering
in comparison to different previous 3DVR methods.
We have been able through this work to raise the quality and efficiency of the neutron
computerized tomography facility in the Egypt Second Research Reactor (ETRR-2) to be able to
study different samples and scientific phenomena with high accuracy and efficiency in showing
the internal structure of the scanned samples in compared to the currently used method in our
facility.