Input Images
λ=0.005
λ=0.010
λ=0.015
λ=0.020
Input Images
λ=0.010
λ=0.020
λ=0.030
λ=0.040
Better smoothing results can be obtained with per-image parameter tweaking.
|
Input Images |
λ=0.005 |
λ=0.010 |
λ=0.015 |
λ=0.020 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Input Images |
λ=0.010 |
λ=0.020 |
λ=0.030 |
λ=0.040 |
|
|
|
|
|
|
|
Input Image |
λ=0.005 |
λ=0.010 |
λ=0.015 |
λ=0.020 |
|
|
|
|
|
The following results are produced by successively applying our algorithm with parameters (λ=2.5e-3, κ=1.05) n times. The bottom row shows the color-coded results.
|
Input Images |
n=1 |
n=5 |
n=10 |
n=20 |
|
|
|
|
|
|
|
|
|
|
Our smoothing method can be regarded as complementary to previous local approaches. New effect can be yielded when successively applying local filtering, e.g., bilateral filtering (BLF), and our smoothing method. Details are in the paper.
|
Input Image |
Ours |
BLF |
BLF+Ours |
|
|
|
|