Title :
Cluster-space hyperspectral data representation for mixed pixel analysis
Author_Institution :
Sch. of Electr. Eng., New South Wales Univ., Campbell, ACT, Australia
Abstract :
The probabilistic mixture model for sub-pixel analysis is investigated and further developed. The probability distribution for each endmember is estimated by using a cluster-space data representation which is adaptable to individual class data distribution shapes. The proposed method is easy to implement and suitable for hyperspectral image data processing. Experiments using HyMap data are presented, showing the detailed sub-pixel estimation
Keywords :
geophysical signal processing; geophysical techniques; multidimensional signal processing; remote sensing; terrain mapping; cluster space data representation; cluster-space hyperspectral data representation; endmember; geophysical measurement technique; hyperspectral remote sensing; image processing; land surface; mixed pixel analysis; multidimensional signal processing; multispectral remote sensing; probabilistic mixture model; probability distribution; sub-pixel analysis; subpixel estimation; terrain mapping; Australia; Educational institutions; Electronic mail; Equations; Gaussian distribution; Hyperspectral imaging; Hyperspectral sensors; Probability distribution; Shape; Training data;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2000. Proceedings. IGARSS 2000. IEEE 2000 International
Conference_Location :
Honolulu, HI
Print_ISBN :
0-7803-6359-0
DOI :
10.1109/IGARSS.2000.858344
