Unsupervised segmentation of synthetic aperture radar sea ice imagery using a novel Markov random field models
Abstract — The operational segmentation of SAR sea ice imagery is a practical, challenging objective in the realm of applied pattern recognition. This research is in support of operational activities at the Canadian Ice Services (CIS), a government agency that monitors all ice-infested regions under...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Text |
Language: | English |
Published: |
2005
|
Subjects: | |
Online Access: | http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.126.1908 http://www.eng.uwaterloo.ca/~dclausi/Papers/Clausi and Deng - PRRS 2004 - MRF Segmentation of SAR Sea Ice Images.pdf |
Summary: | Abstract — The operational segmentation of SAR sea ice imagery is a practical, challenging objective in the realm of applied pattern recognition. This research is in support of operational activities at the Canadian Ice Services (CIS), a government agency that monitors all ice-infested regions under Canadian jurisdiction. This paper uses a fusion of tone and texture to segment SAR sea ice images in an unsupervised manner. A novel Markov random field (MRF) segmentation technique is employed and produces improved results over K-means and the traditional MRF implementation. I. |
---|