TABLE I
Comparison of three fidelity measures and four binary rendition methods, when reproducing 250 noisy versions of Fig. 7. Columns give the percentage of cases for which a given method performed best, second best, third best, or worst, relative to the other methods, under a given fidelity measure
Measure | Spiral | Method | Rank |
---|
Square | Bayer |
---|
MSE | 0 | 0 | 6 | 94 |
11 | 30 | 55 | 6 |
52 | 36 | 17 | 0 |
36 | 34 | 22 | 0 |
C1 | 0 | 0 | 49 | 52 |
0 | 0 | 51 | 48 |
70 | 36 | 0 | 0 |
29 | 64 | 0 | 0 |
C2 | 0 | 0 | 19 | 81 |
28 | 4 | 56 | 18 |
54 | 36 | 12 | 0 |
17 | 60 | 12 | 0 |
TABLE II
Average performance of four binary rendition methods when reproducing 250 noisy versions of Fig. 7
Measure | Spiral | Method | Rank |
---|
Square | Bayer |
---|
MSE | 0.202 | 0.198 | 0.195 | 0.159 |
C1 | 0.205 | 0.198 | 0.272 | 0.275 |
C2 | 0.044 | 0.041 | 0.049 | 0.062 |
TABLE III
Comparison of three fidelity measures and four binary rendition methods, when reproducing 250 white-noise images. Columns give the percentage of cases for which a given method performed best, second best, third best, or worst, relative to the other methods under a given fidelity measure
Measure | Spiral | Method | Rank |
---|
Square | Bayer |
---|
MSE | 0 | 0 | 0 | 100 |
31 | 36 | 35 | 0 |
44 | 35 | 21 | 0 |
25 | 29 | 44 | 0 |
C1 | 4 | 3 | 2 | 91 |
33 | 30 | 28 | 8 |
36 | 39 | 25 | 0 |
27 | 28 | 44 | 0 |
C2 | 0 | 0 | 1 | 99 |
27 | 34 | 38 | 1 |
43 | 42 | 15 | 0 |
30 | 23 | 46 | 0 |
TABLE IV
Average performance of four binary rendition methods when reproducing 250 white-noise images
Measure | Spiral | Methiod | Rank |
---|
Square | Bayer |
---|
MSE | 0.165 | 0.166 | 0.169 | 0.083 |
C1 | 0.339 | 0.338 | 0.334 | 0.373 |
C2 | 0.096 | 0.097 | 0.096 | 0.145 |