Low-Rank Tensor Patching Based on Convolutional Sparse Coding for Communication Data Repair

Table 1

Evaluation results of different algorithms for complementary color images.

Image	MR (%)	PSNR				SSIM
		HaLRTC	TTN	TV	CSC	HaLRTC	TTN	TV	CSC
Airplane 256 × 256 × 3	70	25.21	25.49	26.55	30.05	0.658	0.622	0.913	0.945
	80	22.70	22.91	24.13	27.37	0.541	0.491	0.858	0.913
	90	19.54	19.76	20.96	23.92	0.367	0.291	0.745	0.848
Peppers 256 × 256 × 3	70	24.03	23.54	26.86	31.72	0.934	0.970	0.969	0.986
	80	20.70	20.50	25.54	28.31	0.823	0.950	0.945	0.975
	90	16.92	16.75	20.24	23.96	0.759	0.891	0.871	0.936
Fruits 256 × 256 × 3	70	24.08	24.30	26.81	30.36	0.875	0.864	0.948	0.972
	80	21.65	21.79	24.89	27.59	0.817	0.796	0.921	0.949
	90	18.22	18.40	21.69	23.76	0.716	0.658	0.861	0.901
Starfish 256 × 256 × 3	70	23.68	24.41	26.18	29.78	0.907	0.899	0.959	0.976
	80	20.91	21.17	23.55	26.75	0.848	0.826	0.927	0.958
	90	17.20	17.54	20.19	22.63	0.732	0.680	0.880	0.907
Lena 256 × 256 × 3	70	25.59	25.86	27.69	32.03	0.948	0.946	0.967	0.985
	80	22.82	23.03	25.84	29.00	0.913	0.904	0.951	0.972
	90	19.26	19.47	22.29	25.34	0.845	0.813	0.910	0.943
House 256 × 256 × 3	70	27.30	27.83	28.45	33.00	0.923	0.915	0.956	0.974
	80	24.22	24.57	26.53	30.72	0.861	0.841	0.932	0.960
	90	20.60	20.62	22.50	26.49	0.728	0.664	0.861	0.915