Table 1.

The ASA of tested algorithms on ST images with various noises.

Noise		FCM	FCM_S1	FCM_S2	FCM + GF (ε=0.14)	IFCM_GF (ρ)	FRFCM	MRIFCM_GF (ρ)
3% Gaussian	0.7028		0.9783	0.9742	0.6984	0.7520 (0.047)	0.9982	∗ 0.9993 (0.014)
5% Gaussian	0.6471		0.9166	0.8716	0.6646	0.7166 (0.034)	0.9965	∗ 0.9987 (0.009)
10% Gaussian	0.5806		0.7628	0.7497	0.6172	0.7153 (0.02)	0.9892	∗ 0.9964 (0.008)
15% Gaussian	0.5499		0.7292	0.7139	0.5924	0.7082 (0.017)	0.9425	∗ 0.9907 (0.005)
10% Salt & Pepper	0.9431		0.9389	0.9826	0.9443	∗ 0.9995 (0.008)	0.9991	0.9993 (0.009)
20% Salt & Pepper	0.8873		0.8757	0.9647	0.8916	∗ 0.9993 (0.006)	0.9989	∗ 0.9993 (0.004)
30% Salt & Pepper	0.8304		0.7806	0.9409	0.8366	0.9981 (0.002)	0.9976	∗ 0.9982 (0.003)

^∗The best segmentation accuracy among the group.