|
Stian Normann Anfinsen
|
|
PH.D. STUDENT
Email: stiann(at)phys.uit.no
Research interests:
Automatic clustering, classification and segmentation of POLSAR data
Statistical modeling of POLSAR data
Speckle filtering of POLSAR data
Retrieval of geophysical and biological parameters from POLSAR data
Data fusion of remote sensing data (radar and optical)
Personal homepage
 
|
Research Projects
|
|
Classification of Polarimetric SAR Data
In this project we study general purpose algorithms for classification of PolSAR data, but specific application areas such as land use classification and sea ice classification are also considered. Our main focus is on unsupervised methods. We apply modern algorithms from pattern recognition and machine learning, such as spectral clustering, kernel methods, and information-theoretic learning, in an attempt to improve traditional algorithms such as the Wishart classifier. We also extend the features used in classification from statistics derived from single pixel densities to include statistical information of neighbouring pixels, texture, and information derived from physical modeling and decomposition theory.
[Read more]
|
Publications
|
|
Conference papers
S. N. Anfinsen, R. Jenssen and T. Eltoft, "Spectral Clustering of Polarimetric SAR Data with Wishart-Derived Distance Measures," Proceedings of POLinSAR 2007, Frascati, Italy, January 2007. [paper] [presentation]
A. P. Doulgeris, S. N. Anfinsen and T. Eltoft, "Analysis of Non-Gaussian POLSAR Data," Proceedings of IGARSS '07, Barcelona, Spain, August 2007, to appear.
|
|