Table 5 Comparison of the accuracy measure applied on the RIDB
From: A novel approach to extracting useful information from noisy TFDs using 2D local entropy measures
SNR | Proposed method | Local entropy-based algorithm | RICI TFD threshold | 10% | 15% | 20% |
|---|---|---|---|---|---|---|
Signal 1 | ||||||
−3 dB | Acc= 0.987 | Acc= 0.94 | Acc= 0.885 | Acc= 0.763 | Acc= 0.828 | Acc= 0.849 |
0 dB | Acc= 0.99 | Acc= 0.947 | Acc= 0.91 | Acc= 0.859 | Acc= 0.864 | Acc= 0.856 |
3 dB | Acc= 0.994 | Acc= 0.953 | Acc= 0.939 | Acc= 0.9 | Acc= 0.88 | Acc= 0.87 |
6 dB | Acc= 0.994 | Acc= 0.932 | Acc= 0.993 | Acc= 0.98 | Acc= 0.98 | Acc= 0.98 |
10 dB | Acc= 0.994 | Acc= 0.940 | Acc= 0.993 | Acc= 0.98 | Acc= 0.98 | Acc= 0.98 |
Signal 2 | ||||||
−3 dB | Acc= 0.912 | Acc= 0.918 | Acc= 0.854 | Acc= 0.718 | Acc= 0.793 | Acc= 0.821 |
0 dB | Acc= 0.932 | Acc= 0.926 | Acc= 0.896 | Acc= 0.812 | Acc= 0.848 | Acc= 0.851 |
3 dB | Acc= 0.948 | Acc= 0.936 | Acc= 0.931 | Acc= 0.872 | Acc= 0.871 | Acc= 0.858 |
6 dB | Acc= 0.959 | Acc= 0.932 | Acc= 0.956 | Acc= 0.92 | Acc= 0.92 | Acc= 0.92 |
10 dB | Acc= 0.964 | Acc= 0.938 | Acc= 0.959 | Acc= 0.92 | Acc= 0.92 | Acc= 0.92 |