Berkeley Lab Scientific Computing Seminar

Date:

Wednesday, July 5, 2006

Time:

10:00am-11:00am

Location:

Building 50A-5132

Seminar Speaker:

Kang Zhang
Department of Computer Science
University of Texas at Dallas
http://www.utdallas.edu/~kzhang/

Title:

Multi-level Data Clustering and Visualization with Applications to Geographical Information Systems

Abstract:

The rapidly growing field of Geographical Information Systems (GIS) integrates data (such as satellite observations), information, geography, etc. into coherent systems. In this talk we focus on the applications of a data clustering framework called FACADE (Fast and Automatic Clustering Approach to Data Engineering). Three features of FACADE distinguish it from other clustering approaches: highly effective noise removal, data compression, and visual support for human participation. FACADE has been applied to GIS remote sensing images and other image processing problems. I will specifically talk about spatial noise removal for post-classification smoothing of GIS images and will give a demo of FACADE (details at http://viscomp.utdallas.edu/FACADE).

Sponsor of Seminar:

Chris Ding

Scientific Computing

Contact Esmond G. Ng EGNg@lbl.gov