Mining Solar Images to support Astrophysics Research

1
Mining Solar Images to support Astrophysics
Research

• Olfa Nasraoui
• olfa.nasraoui_at_louisville.edu
• Project Summary
• This projects goals are to develop an image
  retrieval system based on Data Mining to quickly
  sift through data sets downloaded from online
  solar image databases and automatically discover
  the rare but interesting images containing solar
  loops
• Plans for Upcoming Year
• Investigate more complex features to use in 2nd
  stage classifier
• Integrate Hough space analysis with stream
  clustering
• Investigate interactions between solar cycle and
  features classifiers
• Continue with in-disk loops and TRACE data.

• FY06 Highlights
• Developed tools to aid in marking images with
  loop locations and feature extraction
• Studied several classifiers within a 2-stage
  framework
• simpler features used in stage 1,
• followed by more costly features in a 2nd stage
  applied only to image blocks that are detected in
  previous stage
• Developed scalable clustering algorithm to
  extract clusters/peaks from Hough Space

• Applications
• Scientific analysis for the Coronal Heating
  problem requires data observed by instruments
  such as EIT, TRACE, and SXT to measure the
  temperature distribution along the coronal loop
  length and study different temperature models
• The manual search for interesting images (with
  coronal loops) is by far the most time consuming
  aspect of this analysis.
• The tools developed as part of this project will
  automate this process

Data mining, coronal loops
NASA AISR 2004-07
http//webmining.spd.louisville.edu/solar