Hyperspectral image classification with SVM and guided filter

EURASIP Journal on Wireless Communications and Networking

Table 2 Classification accuracy on the University of Pavia dataset (10% samples for training)

	SVM	SVM-EPF	Co-SVM	Co-SVM-EPF	GF-SVM	GF-SVM-EPF
Asphalt	98.05	99.37	99.46	98.68	99.06	99.61
Meadows	98.26	99.69	99.96	99.80	99.91	99.87
Gravel	79.34	95.21	98.03	99.94	96.74	100.0
Trees	92.93	99.68	99.15	99.67	97.93	98.60
P-M-sheets	94.44	98.75	100.0	99.54	100.0	100.0
Bare Soil	84.18	96.75	99.74	99.67	99.41	99.76
Bitumen	72.48	99.65	99.65	100.0	99.88	100.0
S-B-Bricks	87.38	93.22	95.83	95.70	97.56	99.31
Shadows	99.79	99.79	100.0	99.57	99.79	99.79
OA	93.36	98.51	99.36	99.24	99.25	99.70
AA	89.65	98.01	99.09	99.17	98.92	99.66
KA	90.92	97.94	99.11	98.95	98.97	99.58