The table below is a complete set of results for demosaicking using the Bayer structure. It provides the 24 original and demosaicked images in TIFF format (uncompressed) and in JPG format (saved with ‘quality’ 90 in MATLAB). Each cell shows the CPSNR (dB) and the S-CIELAB metric.
The methods used to obtain the results are described below the table.
| No. | Thumb | Original | A | B | C | D | E | F | G (ST) | H (TO) | I (STRE) | J (TORE) |
| 1 |  |
TIFF | TIFF
37.79 1.255 |
TIFF
36.08 1.355 |
TIFF
38.51 1.108 |
TIFF
34.98 1.248 |
TIFF
38.28 1.198 |
TIFF
38.13 1.156 |
TIFF
39.55 1.052 |
TIFF
38.67 1.133 |
TIFF
38.17 1.193 |
TIFF
37.92 1.190 |
| 2 |  |
TIFF | TIFF
39.64 0.719 |
TIFF
40.14 0.704 |
TIFF
40.92 0.648 |
TIFF
39.11 0.801 |
TIFF
39.84 0.718 |
TIFF
40.01 0.671 |
TIFF
41.54 0.581 |
TIFF
40.81 0.642 |
TIFF
40.84 0.628 |
TIFF
40.65 0.664 |
| 3 |  |
TIFF | TIFF
41.61 0.556 |
TIFF
41.88 0.601 |
TIFF
42.73 0.479 |
TIFF
40.91 0.574 |
TIFF
42.10 0.548 |
TIFF
41.65 0.534 |
TIFF
43.22 0.481 |
TIFF
43.37 0.509 |
TIFF
42.76 0.490 |
TIFF
42.22 0.512 |
| 4 |  |
TIFF | TIFF
40.16 0.727 |
TIFF
40.45 0.76 |
TIFF
41.08 0.671 |
TIFF
38.73 0.851 |
TIFF
40.85 0.706 |
TIFF
40.62 0.688 |
TIFF
42.01 0.610 |
TIFF
41.34 0.644 |
TIFF
41.38 0.639 |
TIFF
41.09 0.666 |
| 5 |  |
TIFF | TIFF
37.57 1.097 |
TIFF
36.75 1.170 |
TIFF
38.09 1.070 |
TIFF
35.26 1.309 |
TIFF
38.01 1.117 |
TIFF
37.85 1.052 |
TIFF
38.78 0.941 |
TIFF
38.30 1.024 |
TIFF
38.09 1.021 |
TIFF
37.92 1.071 |
| 6 |  |
TIFF | TIFF
38.62 0.950 |
TIFF
38.03 1.004 |
TIFF
40.25 0.774 |
TIFF
37.36 0.864 |
TIFF
39.84 0.851 |
TIFF
40.03 0.777 |
TIFF
41.01 0.721 |
TIFF
40.42 0.771 |
TIFF
40.04 0.794 |
TIFF
39.95 0.784 |
| 7 |  |
TIFF | TIFF
41.83 0.63 |
TIFF
41.51 0.692 |
TIFF
42.38 0.574 |
TIFF
40.36 0.686 |
TIFF
42.43 0.577 |
TIFF
42.12 0.537 |
TIFF
43.59 0.468 |
TIFF
42.86 0.526 |
TIFF
42.87 0.533 |
TIFF
42.52 0.556 |
| 8 |  |
TIFF | TIFF
35.15 1.548 |
TIFF
34.35 1.529 |
TIFF
36.06 1.332 |
TIFF
33.60 1.434 |
TIFF
36.02 1.400 |
TIFF
35.22 1.482 |
TIFF
36.16 1.433 |
TIFF
35.66 1.449 |
TIFF
35.45 1.561 |
TIFF
35.32 1.475 |
| 9 |  |
TIFF | TIFF
41.85 0.640 |
TIFF
41.70 0.681 |
TIFF
42.84 0.569 |
TIFF
40.71 0.649 |
TIFF
42.07 0.648 |
TIFF
42.06 0.591 |
TIFF
43.05 0.569 |
TIFF
42.85 0.581 |
TIFF
42.61 0.583 |
TIFF
42.57 0.585 |
| 10 |  |
TIFF | TIFF
42.04 0.587 |
TIFF
41.55 0.648 |
TIFF
42.61 0.544 |
TIFF
40.45 0.620 |
TIFF
42.47 0.601 |
TIFF
42.21 0.549 |
TIFF
43.00 0.532 |
TIFF
42.85 0.541 |
TIFF
42.70 0.543 |
TIFF
42.63 0.547 |
| 11 |  |
TIFF | TIFF
39.25 0.844 |
TIFF
38.32 0.890 |
TIFF
40.08 0.737 |
TIFF
37.25 0.856 |
TIFF
39.70 0.800 |
TIFF
39.74 0.759 |
TIFF
40.43 0.725 |
TIFF
40.33 0.731 |
TIFF
39.85 0.760 |
TIFF
39.81 0.758 |
| 12 |  |
TIFF | TIFF
42.63 0.530 |
TIFF
42.22 0.611 |
TIFF
43.51 0.461 |
TIFF
41.53 0.520 |
TIFF
43.19 0.519 |
TIFF
43.05 0.464 |
TIFF
43.72 0.472 |
TIFF
43.46 0.479 |
TIFF
43.23 0.482 |
TIFF
43.14 0.488 |
| 13 |  |
TIFF | TIFF
34.37 1.730 |
TIFF
32.38 1.913 |
TIFF
34.80 1.560 |
TIFF
31.02 1.943 |
TIFF
34.84 1.625 |
TIFF
35.07 1.583 |
TIFF
36.57 1.421 |
TIFF
34.89 1.573 |
TIFF
34.69 1.804 |
TIFF
34.20 1.629 |
| 14 |  |
TIFF | TIFF
35.82 1.143 |
TIFF
36.69 1.108 |
TIFF
37.01 0.981 |
TIFF
35.12 1.202 |
TIFF
36.10 1.079 |
TIFF
35.85 1.079 |
TIFF
37.82 0.923 |
TIFF
36.75 0.991 |
TIFF
37.25 0.955 |
TIFF
36.77 1.001 |
| 15 |  |
TIFF | TIFF
39.42 0.686 |
TIFF
39.26 0.738 |
TIFF
39.87 0.657 |
TIFF
37.57 0.798 |
TIFF
39.89 0.685 |
TIFF
39.67 0.659 |
TIFF
40.67 0.584 |
TIFF
40.03 0.636 |
TIFF
40.06 0.618 |
TIFF
39.83 0.648 |
| 16 |  |
TIFF | TIFF
41.88 0.675 |
TIFF
41.60 0.714 |
TIFF
43.81 0.531 |
TIFF
41.22 0.587 |
TIFF
43.24 0.593 |
TIFF
43.66 0.531 |
TIFF
44.45 0.497 |
TIFF
44.09 0.526 |
TIFF
43.68 0.530 |
TIFF
43.63 0.537 |
| 17 |  |
TIFF | TIFF
41.44 0.555 |
TIFF
40.20 0.633 |
TIFF
41.84 0.517 |
TIFF
38.93 0.612 |
TIFF
41.49 0.559 |
TIFF
41.57 0.524 |
TIFF
42.17 0.501 |
TIFF
42.00 0.507 |
TIFF
41.70 0.534 |
TIFF
41.63 0.517 |
| 18 |  |
TIFF | TIFF
37.35 1.069 |
TIFF
36.03 1.106 |
TIFF
37.47 1.008 |
TIFF
34.39 1.290 |
TIFF
37.47 1.033 |
TIFF
37.44 1.022 |
TIFF
37.89 0.983 |
TIFF
37.80 0.965 |
TIFF
37.34 1.023 |
TIFF
37.28 0.997 |
| 19 |  |
TIFF | TIFF
39.76 0.837 |
TIFF
39.20 0.890 |
TIFF
40.90 0.752 |
TIFF
38.18 0.874 |
TIFF
40.01 0.844 |
TIFF
40.35 0.790 |
TIFF
41.04 0.735 |
TIFF
40.87 0.755 |
TIFF
40.56 0.771 |
TIFF
40.53 0.771 |
| 20 |  |
TIFF | TIFF
40.72 0.592 |
TIFF
39.90 0.679 |
TIFF
41.27 0.545 |
TIFF
38.88 0.642 |
TIFF
41.00 0.594 |
TIFF
40.39 0.601 |
TIFF
41.44 0.524 |
TIFF
41.25 0.567 |
TIFF
41.05 0.542 |
TIFF
40.97 0.578 |
| 21 |  |
TIFF | TIFF
38.77 0.960 |
TIFF
37.16 1.060 |
TIFF
39.16 0.901 |
TIFF
36.32 1.030 |
TIFF
39.22 0.940 |
TIFF
38.74 0.935 |
TIFF
39.65 0.868 |
TIFF
39.28 0.892 |
TIFF
38.91 0.931 |
TIFF
38.79 0.917 |
| 22 |  |
TIFF | TIFF
37.71 1.057 |
TIFF
37.60 1.054 |
TIFF
38.44 0.964 |
TIFF
36.17 1.215 |
TIFF
38.29 0.972 |
TIFF
38.09 0.940 |
TIFF
38.86 0.884 |
TIFF
38.67 0.908 |
TIFF
38.61 0.914 |
TIFF
38.57 0.930 |
| 23 |  |
TIFF | TIFF
42.01 0.553 |
TIFF
42.53 0.604 |
TIFF
43.28 0.495 |
TIFF
41.46 0.614 |
TIFF
42.35 0.536 |
TIFF
42.12 0.506 |
TIFF
44.25 0.440 |
TIFF
43.15 0.479 |
TIFF
43.67 0.461 |
TIFF
42.98 0.497 |
| 24 |  |
TIFF | TIFF
34.75 1.132 |
TIFF
34.28 1.179 |
TIFF
35.53 1.018 |
TIFF
32.47 1.312 |
TIFF
35.57 1.055 |
TIFF
35.39 1.012 |
TIFF
35.75 0.990 |
TIFF
35.60 0.969 |
TIFF
35.37 1.043 |
TIFF
35.34 0.989 |
The demosaicking methods used in the table are the following:
Method A: Bayer_results/POCS (B. K. Gunturk, Y. Altunbasak, and R. M. Mersereau, “Color plane interpolation using alternating projections,” IEEE Trans. Image Process., vol. 11, no. 9, pp. 997–1013, Sep. 2002.)
Method B: Bayer_results/PCSD (X. Wu and N. Zhang, “Primary-consistent soft-decision color demosaicking for digital cameras (patent pending),” IEEE Trans. Image Process., vol. 13, no. 9, pp. 1263–1274, Sep. 2004.)
Method C: Bayer_results/LMMSE (L. Zhang and X. Wu, “Color demosaicking via directional linear minimum mean square-error estimation,” IEEE Trans. Image Process., vol. 14, no. 12, pp. 2167–2177, Dec. 2005.)
Method D: Bayer_results/AHD (K. Hirakawa and T. W. Parks, “Adaptive homogeneity-directed demosaicing algorithm,” IEEE Trans. Image Process., vol. 14, no. 3, pp. 360–369, Mar. 2005.)
Method E: RAD (D. Menon and G. Calvagno, “Regularization approaches to demosaicking,” IEEE Trans. Image Process., vol. 18, no. 10, pp. 2209–2220, Oct. 2009.)
Method F: Bayer_results/FDM (E. Dubois, “Frequency-domain methods for demosaicking of Bayersampled color image,” IEEE Signal Process. Lett., vol. 12, pp. 847–850, Dec. 2005.)
Method G: LSLCD with ST filter (configuration [11 11 11 11 11 11], obtained by self-training on the given image)
Method H: LSLCD with TO filter (configuration [11 11 11 11 11 11], obtained by training on other 23 images)
Method I: LSLCD with RE filter (configuration [5 5 9 3 11 1], obtained by self-training on the given image)
Method J: LSLCD with RE filter (configuration [5 5 9 3 11 1], obtained by training on other 23 images)
