Sparse Regularization Based on Orthogonal Tensor Dictionary Learning for Inverse Problems

Table 1

SNR values of three different methods applied to synthetic data in Figure 2 with different sampling ratio.

Sampling ratio	0.3	0.4	0.5	0.6	0.7	0.8
K-SVD method	23.9923	25.6434	27.8589	29.3336	30.6696	31.1826
Orthogonal DL method	31.7795	32.9216	33.4244	36.3080	38.4702	41.3320
Proposed method	34.6241	36.3963	38.5860	39.9641	42.9120	43.5903