Title :
A Fuzzy Homogeneity Test for the Iterative Guided Spectral Class Rejection Algorithm
Author :
Phillips, Rhonda D. ; Watson, Layne T. ; Wynne, Randolph H.
Author_Institution :
Depts. of Comput. Sci. & Math., Virginia Polytech. Inst. & State Univ., Blacksburg, VA
Abstract :
This work is an intermediate step toward a fuzzy version of the iterative guided spectral class rejection (IGSCR) classification algorithm. IGSCR combines a clustering algorithm and a decision rule to produce multiple classifications. Although fuzzy versions of these algorithms are available, IGSCR can only evaluate hard clustering output. In an effort to move toward fuzzy classification output, this paper presents a statistically rigorous fuzzy cluster homogeneity test that is analogous to the discrete cluster homogeneity test.
Keywords :
decision theory; fuzzy logic; geophysical techniques; geophysics computing; image classification; pattern clustering; IGSCR classification algorithm; clustering algorithm; decision rule; fuzzy homogeneity test; hard clustering output; iterative guided spectral class rejection algorithm; multiple classifications; Classification algorithms; Clustering algorithms; Clustering methods; Computer science; Fuzzy sets; Iterative algorithms; Partitioning algorithms; Pixel; Probability; Testing; classification; clustering; fuzzy clustering; hypothesis testing;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2008. IGARSS 2008. IEEE International
Conference_Location :
Boston, MA
Print_ISBN :
978-1-4244-2807-6
Electronic_ISBN :
978-1-4244-2808-3
DOI :
10.1109/IGARSS.2008.4779136
