A New Methodology for Spectral-Spatial Classification of Hyperspectral Images

Table 4

Comparison of class-specific accuracies in percentage for the Indiana image by different methods.

Class	Samples		Method
	Train	Test	Pixel-wise SVM (%)	Three PCs (%)	(%)	(%)	(%)
C1	422	1012	84.67	88.91	92.12	93.10	93.51
C2	252	582	73.91	72.54	93.65	94.96	95.20
C3	392	902	95.79	97.76	99.92	98.38	98.92
C4	150	347	92.70	93.36	95.98	96.38	96.38
C5	198	416	83.91	97.88	98.86	98.53	99.19
C6	232	515	96.39	89.96	98.53	99.33	99.06
C7	150	339	99.40	99.18	99.18	99.39	99.39
C8	277	691	71.53	75.83	78.41	78.10	78.31
C9	52	160	99.30	99.06	99.53	100	100
C10	64	170	64.44	87.61	75.64	79.06	78.21
C11	112	268	73.99	78.16	99.47	96.05	96.05
C12	760	1708	85.00	96.56	98.91	98.78	98.99