|
Computer vision and Artificial
Intelligence techniques are now widely used in a variety of
industrial and scientific fields. My interest in the use of these tools
is to design and construct robust automatic galaxy classifiers. This work
will incorporate knowledge from advance image processing and artificial
neural networks.
Why this research interest?
Classification of galaxies is an important issue in the large scale study
of the Universe. The recent release of the Hubble Deep Field (HDF) image and
initiation of several large scale sky surveys demonstrate that our
universe is vast and not empty. According to estimates computed from the
HDF, astronomers predict that the universe may potentially contain over
50 billion galaxies. Database construction of these galaxies has already
began. Recognition and classification of this number of galaxies is not a
simple manual task. Efficient and robust computer automated classifiers have
to be investigated and researched to help astronomers derive the maximum
benefit form survey studies. My goal in building these automated
classifiers is to apply newly discovered techniques in computer vision
and artificial intelligence to recognition and classification of galaxies
and the analysis of a large number of fields.
The role of students:
This type of research is in the frontiers of physics, as such students
can acquire skills necessary to secure jobs in toady's market. The skills
are learning about image processing, working with advance image
processing tools like KHOROS, design, implementation and testing of
algorithms in C/C++, learning what is artificial
neural networks (ANNS) and a variety
of software tools used in today's science. Most of all they learn
how to use these tools in physics and astronomy.
Equipment and facility:
These days PC computers are fast and are also relatively inexpensive.
Currently I am using a PC computer with Linux operating system as a Unix
workstation in my office. KHOROS and other imaging tools reside on this
workstation. I also use the physics computer laboratory in the department
of physics. Artificial neural network software (Stuttgart Neural Network Simulator
SNNS), can be accessed on the entropy server in this laboratory. A
scanner and postscript printers are also available in this laboratory.
If your interest are similar, or would like
to work with me in collaboration then drop me an email.
Thank you for visiting here.
|
