Input Image
 |
Lambda=5 (RGB) | Lambda=5 (DM, t=4) |
 |
 |
|
| Detail enhancement using RGB smoothing | Detail enhancement using DM smoothing |
|
 |
 |
Here we compare the use of RGB and diffusion distances for edge-aware smoothing using the weighted least-squares method. The images below show that the strong edges at the skyline are better preserved using the diffusion maps while the same overall amount of smoothing is applied. When this smoothing is used for detail enhancement, a more uniform enhancement is obtained by the diffusing distances. This effect is most noticeable at the upper part of the two building on the left.
Instructions: Roll your mouse over each thumbnail to see a larger version of the image appear below. This makes it possible to easily flip back and forth between the images. Click on a thumbnail for a full resolution version of the image.
Input Image Lambda=5 (RGB) Lambda=5 (DM, t=4) Detail enhancement using RGB smoothing Detail enhancement using DM smoothing