EURASIP Journal on Advances in Signal Processing

Table 5 Average SSIM scores by different methods

From: Image denoising based on global image similar patches searching and HOSVD to patches tensor

\(\sigma\)	20		40		60
	SSIM	NCC	SSIM	NCC	SSIM	NCC
BM3D [13]	0.873	0.934	0.792	0.836	0.730	0.798
\(BM3D_{bst}\) [20]	0.882	0.951	0.798	0.841	0.736	0.807
EPLL [17]	0.869	0.917	0.785	0.821	0.723	0.780
WNNM [31]	0.884	0.952	0.793	0.836	0.734	0.801
SLR [23]	0.878	0.938	0.792	0.836	0.731	0.798
LR-GSC [32]	0.883	0.951	0.802	0.853	0.734	0.801
Ours	0.881	0.951	0.797	0.841	0.738	0.813

Bold value indicates the best SSIM and NCC results for all of the methds (the first column)

Back to article page