Table 2.

AREs for four key physiological points of three types of tactile sensors after eight denoising methods.

		Butterworth [19]	Wavelet [20]	EMD [22]	VMD [24]	MA [26]	RPCA [27]	WRPCA [28]	cc-DWRPCA [13]	ANOVAa
3 × 4	P1	0.068 ± 0.029	0.093 ± 0.024	0.100 ± 0.030	0.071 ± 0.032	0.068 ± 0.024	0.197 ± 0.016	0.067 ± 0.033	0.062 ± 0.014	p < 0.05
	P2	0.141 ± 0.062	0.198 ± 0.066	0.154 ± 0.065	0.145 ± 0.065	0.170 ± 0.078	0.174 ± 0.024	0.143 ± 0.067	0.076 ± 0.020	p < 0.05
	P3	0.249 ± 0.134	0.378 ± 0.125	0.276 ± 0.141	0.249 ± 0.137	0.259 ± 0.132	0.166 ± 0.035	0.253 ± 0.147	0.139 ± 0.078	p < 0.05
	P4	0.168 ± 0.081	0.187 ± 0.083	0.172 ± 0.080	0.176 ± 0.085	0.159 ± 0.081	0.154 ± 0.018	0.143 ± 0.077	0.063 ± 0.034	p < 0.05
4 × 6	P1	0.239 ± 0.216	0.244 ± 0.202	0.270 ± 0.199	0.243 ± 0.215	0.230 ± 0.196	0.279 ± 0.084	0.258 ± 0.242	0.174 ± 0.144	p < 0.05
	P2	0.526 ± 0.323	0.583 ± 0.304	0.527 ± 0.322	0.529 ± 0.320	0.573 ± 0.347	0.307 ± 0.195	0.560 ± 0.392	0.258 ± 0.322	p < 0.05
	P3	0.934 ± 0.514	1.096 ± 0.507	0.939 ± 0.512	0.923 ± 0.512	0.952 ± 0.510	0.425 ± 0.331	0.939 ± 0.501	0.457 ± 0.410	p < 0.05
	P4	0.653 ± 0.429	0.645 ± 0.440	0.651 ± 0.422	0.662 ± 0.425	0.631 ± 0.426	0.325 ± 0.220	0.596 ± 0.362	0.258 ± 0.270	p < 0.05
5 × 5−1	P1	0.030 ± 0.023	0.077 ± 0.019	0.174 ± 0.068	0.078 ± 0.136	0.041 ± 0.019	0.246 ± 0.030	0.039 ± 0.025	0.038 ± 0.016	p < 0.05
	P2	0.060 ± 0.031	0.156 ± 0.043	0.124 ± 0.054	0.128 ± 0.214	0.077 ± 0.047	0.249 ± 0.037	0.087 ± 0.036	0.065 ± 0.012	p < 0.05
	P3	0.095 ± 0.057	0.258 ± 0.075	0.224 ± 0.129	0.222 ± 0.357	0.103 ± 0.064	0.224 ± 0.047	0.109 ± 0.064	0.074 ± 0.041	p < 0.05
	P4	0.074 ± 0.045	0.142 ± 0.039	0.096 ± 0.046	0.189 ± 0.307	0.070 ± 0.039	0.238 ± 0.035	0.063 ± 0.041	0.044 ± 0.019	p < 0.05

^aThe one-way ANOVA results indicate the significant difference among AREs of the eight algorithms.