link to main page

LSLCD Algorithm Performance on Different Sampling Structures

The table lists the CPNSR value obtained on each of the 24 Kodak images. For each image, the training set used is the other 23 images of the Kodak set. The filters used in this comparison have a 11x11 size. For each image we indicate the best result in bold.

In general, the images sampled with a Bayer structure give a better objective demosaicking performance than the ones sampled with the diagonal stripe structure.

Looking at the images from a subjective point of view:

No. Thumb Bayer Stripe
1 thumb 1 38.67 38.81
2 thumb 2 40.81 38.93
3 thumb 3 43.37 39.46
4 thumb 4 41.34 39.59
5 thumb 5 38.30 35.31
6 thumb 6 40.42 38.96
7 thumb 7 42.86 40.43
8 thumb 8 35.66 36.48
9 thumb 9 42.85 40.99
10 thumb 10 42.85 41.10
11 thumb 11 40.33 38.68
12 thumb 12 43.46 42.02
13 thumb 13 34.89 32.91
14 thumb 14 36.75 34.63
15 thumb 15 40.03 38.88
16 thumb 16 44.09 42.56
17 thumb 17 42.00 40.34
18 thumb 18 37.80 36.05
19 thumb 19 40.97 39.95
20 thumb 20 41.25 39.67
21 thumb 21 39.28 38.82
22 thumb 22 38.67 37.36
23 thumb 23 43.15 40.89
24 thumb 24 35.60 34.17