Sparse Representation Classification Based on Flexible Patches Sampling of Superpixels for Hyperspectral Images

Table 4

The classification accuracy (average on ten runs with 1% random samples) for Salinas images on six test approaches.

class	PSRC	JSRM	MJSR	MASR	BTC-WSL	SRC-FPSS
C1	98.11	100.00	100.00	100.00	100.00	100.00
C2	98.47	99.83	99.89	99.97	100.00	100.00
C3	96.32	98.72	99.95	99.89	100.00	98.77
C4	98.34	98.64	99.32	93.66	99.92	95.92
C5	96.97	99.25	99.10	94.06	99.92	99.29
C6	99.73	100.00	100.00	97.98	100.00	99.92
C7	99.32	99.94	99.88	97.97	100.00	100.00
C8	75.82	88.07	95.07	99.43	100.00	99.69
C9	98.73	99.97	99.78	100.00	100.00	100.00
C10	92.23	97.98	98.46	98.14	100.00	99.71
C11	98.42	99.51	100.00	99.11	100.00	99.70
C12	99.40	99.95	99.95	95.90	100.00	99.89
C13	96.90	99.20	99.89	92.99	99.15	99.31
C14	95.87	98.82	99.51	95.96	99.48	99.51
C15	64.53	79.98	91.93	99.15	88.75	98.70
C16	98.66	99.36	99.71	99.36	100.00	98.78
OA	88.79	94.50	97.67	98.46	98.53	99.48
AA	94.24	9745	98.90	99.20	97.72	99.32
Kappa	85.42	93.87	97.40	98.28	98.36	99.42