Distributed Lunar Image Analysis System
Senior Project: 2004-2005
Patrick Clary, Micah Dowty, Jason Patterson, Christine Predaina and Douglas Wendel
With NASA's new directive to focus on Mars and Moon exploration, it has become
imperative to analyze the countless gigabytes of information previously
gathered about these two celestial bodies. There exist hundreds of thousands of
photos taken of both the Moon and Mars that have yet to be analyzed. NASA has
an active outreach program aimed at involving the general public in space
sciences in an effort to instill excitement about their programs. The challenge
was to develop a publicly accessible web site where these images can be
analyzed by anyone interested in space, geology, online games and puzzles, or
perhaps just for adventure and amusement.
Similar to the SETI@HOME
project where thousands of home computers are used for their processing power,
this project provides a website to harness the "brain power" of thousands of
public users. By enlisting a large pool of public users to view the images in
the data set, large numbers of images can be processed without requiring the
time and resources of the professional scientific community. By combining the
individual analysis results, the system is able to determine which images
contain features of interest.
The project was implemented as a three-tier web application. The bottom tier is
a database engine, currently being hosted on a MySQL database server. The middle
tier is managed by a C#/.NET layer running on the Mono application server. The
top tier is presented as HTML and CSS to the end user via a supported web
browser. By utilizing this architecture, the system is both flexible and
modular, allowing entities within each tier to be changed without drastically
affecting the connective layers. The system was designed to be easily
extensible to support analysis of images in a wide variety of disciplines.



|