Spectral-Spatial Hyperspectral Image Classification Based on Homogeneous Minimum Spanning Forest

Table 3

Information classes, number of training samples, and classification accuracy for Indian Pines.

	No. of samples	No. of training samples	SVM	MC-MSF	RD-MSF	MSWMF-MSF	MSMMSF	Proposed Methods
								HMSF	HKNN-MSF-MV	F–HKNN—MSF-MV
Alfalfa	46	15	95.65	94.87	97.83	97.83	100	97.83	100	100
Corn-notill	1428	50	64.85	75.13	78.06	93.77	86.55	62.18	88.17	96.5
Corn-mintill	830	50	61.57	74.62	86.99	97.11	90.72	67.23	79.76	94.82
Corn	237	50	86.08	92.39	91.85	96.62	100	96.2	96.2	100
Grass-pasture	483	50	92.55	98.21	98.7	98.34	94.62	88.8	95.24	97.52
Grass-trees	730	50	87.81	93.51	92.17	98.49	97.95	98.49	98.9	99.86
Grass-p-mowed	28	15	96.43	96.43	96.43	96.43	100	100	96.43	96.43
Hay-windrowed	478	50	99.16	99.54	99.77	99.79	100	100	99.75	99.79
Oats	20	15	100	100	100	100	100	100	95	100
Soybean-notill	972	50	66.46	81.68	88.14	94.24	88.79	74.59	84.1	90.53
Soybean-mintill	2455	50	47.94	64.14	68.82	85.24	85.38	60.33	63.1	83.83
Soybean-clean	593	50	71.5	85.99	90.78	96.46	92.22	85.5	86.68	98.48
Wheat	205	50	99.02	99.38	99.38	99.51	100	99.02	99.5	99.51
Woods	1265	50	83.32	88.38	88.85	98.74	97.15	88.06	93.36	97.94
B-G-T-drives	386	50	71.76	82.12	81.52	.98.7	98.96	96.11	91.45	98.19
S-steel-towers	93	50	91.4	97.78	97.78	98.92	100	98.92	100	100
AA (%)	—	—	82.21	89.06	91.06	96.93	95.64	88.33	91.73	97.08
OA (%)	—	—	69.79	82.14	87.17	93.15	91.45	76.94	84.03	93.76
Kappa coefficient	—	—	66.27	80.56	85.78		90.32	74.2	82.05	92.92