Title :
Unsupervised classification using spatial region growing segmentation and fuzzy training
Author :
Lee, Sanghoon ; Crawford, Melba M.
Author_Institution :
Dept. of Ind. Eng., Kyungwon Univ., Kyunggi-Do, South Korea
Abstract :
This study has utilized the approach to unsupervisedly estimate the number of classes and the parameters of defining the classes in order to train the classifier. A region growing segmentation and local fuzzy classification have been employed to rind the sample classes that well represent the ground truth. The maximum likelihood classifier has then used the sample classes
Keywords :
geophysical signal processing; geophysical techniques; image classification; image segmentation; remote sensing; terrain mapping; fuzzy training; geophysical measurement technique; image classification; image processing; image segmentation; land surface; maximum likelihood classifier; remote sensing; spatial region growing; terrain mapping; training; unsupervised classification; Cams; Clustering algorithms; Computational efficiency; Image processing; Image segmentation; Layout; Merging; Optical sensors; Partitioning algorithms; Pixel;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2001. IGARSS '01. IEEE 2001 International
Conference_Location :
Sydney, NSW
Print_ISBN :
0-7803-7031-7
DOI :
10.1109/IGARSS.2001.978195
