Description:
The overall goal of this project is to create time-efficient mechanisms for classifying and querying multiple types of solar phenomena. There are several steps in our approach. The first step is to extract informative features from solar images. The second step is to build accurate classifiers for the solar phenomena. Finally, we plan to build an interface that allows querying of TRACE Solar Data repository, which uses the classifiers that have been produced. We are focused on development of querying mechanisms that allow the user to search through solar images based on the image of phenomena, in which he/she is interested.
Support:
Currently, the work is supported by subcontract with Lockheed-Martin in Palo Alto, which has funds from NASA's TRACE Mission. The research is conducted in close collaboration with Dr. Petrus C. Martens, research professor at the MSU, who specializes in area of solar physics.
Opportunities:
Opportunities exist for graduate and undergraduate Computer Science students who are interested in feature recognition and image processing.
2008-09-03