A study of fuzzy clustering within the IGSCR framework

Authors:
Rhonda D. Phillips;Layne T. Watson;Randolph H. Wynne
Affiliations:
Virginia Polytechnic Institute and State University, Blacksburg, VA;Virginia Polytechnic Institute and State University, Blacksburg, VA;Virginia Polytechnic Institute and State University, Blacksburg, VA
Venue:
Proceedings of the 46th Annual Southeast Regional Conference on XX
Year:
2008

Citing 4
Cited 0

Introductory Digital Image Processing: A Remote Sensing Perspective

Introductory Digital Image Processing: A Remote Sensing Perspective
Clustering Algorithms

Clustering Algorithms
Hybrid image classification and parameter selection using a shared memory parallel algorithm

Computers & Geosciences
Data Clustering: Theory, Algorithms, and Applications (ASA-SIAM Series on Statistics and Applied Probability)

Data Clustering: Theory, Algorithms, and Applications (ASA-SIAM Series on Statistics and Applied Probability)

Quantified Score

Hi-index	0.00

Visualization

Abstract

The iterative guided spectral class rejection (IGSCR) classification algorithm uses an underlying clustering method and a decision rule to arrive at final classifications for remotely sensed data. Previous versions of IGSCR have used a hard clustering method such as k-means or ISODATA. In an effort to ultimately create a fuzzy version of IGSCR, this work uses an underlying fuzzy clustering algorithm within the IGSCR framework to study the effects of using the fuzzy clustering algorithm. IGSCR with fuzzy k-means was applied to a Landsat ETM+ satellite image to produce a two class classification (forest and nonforest), and results show that although fuzzy k-means did not lead to increased accuracy, the classification results are dramatically different for IGSCR using traditional k-means and fuzzy k-means.