A Research on the Combination Strategies of Multiple Features for Hyperspectral Remote Sensing Image Classification

Table 5

Experimental results of different datasets associated with different scenarios. “Sc.” represents different scenarios in Table 4.

Method	IP				SL				PU
	Sc.1	Sc.2	Sc.3	Sc.4	Sc.1	Sc.2	Sc.3	Sc.4	Sc.1	Sc.2	Sc.3	Sc.4
PC	81.31	93.60	94.31	96.08	89.52	95.9	98.28	98.44	82.87	85.7	97.44	98.67
Gabor	0.784	0.927	0.935	0.955	0.833	0.954	0.981	0.983	0.765	0.807	0.966	0.983
PC	81.31	48.19	78.52	77.25	89.52	57.16	89.69	89.72	82.87	65.83	91.17	87.36
GLCM	0.784	0.365	0.752	0.737	0.833	0.508	0.885	0.885	0.765	0.503	0.882	0.83
LE	81.31	93.60	80.11	95.66	89.52	95.9	95.51	98.78	82.87	85.7	96.17	97.26
Gabor	0.784	0.927	0.772	0.95	0.833	0.954	0.95	0.986	0.765	0.807	0.949	0.964
LE	81.31	48.19	61.92	68.02	89.52	57.16	83.66	85.34	82.87	65.83	85.51	81.98
GLCM	0.784	0.561	0.561	0.628	0.833	0.508	0.817	0.837	0.765	0.503	0.805	0.754