 |
The 400 x 300 pixel RGB color image (a photograph taken by myself during a visit to Barcelona, Spain) used as source function in the approximation examples below. The image was sampled at 1,600 random positions, and reconstructed using several "classic" scattered data methods. |
 |
Reconstruction using Shepard's global method. |
 |
Reconstruction using Shepard's local method, using one closest sample. In this case, Shepard's local method degenerates to a piecewise constant-valued Voronoi diagram of the sample points. |
 |
Reconstruction using Shepard's local method, using 10 closest samples. |
 |
Reconstruction using Shepard's local method, using 100 closest samples. |
 |
Reconstruction using Hardy's global method, using parameters R = 1.0 and a = 0.5. |
 |
Reconstruction using Hardy's global method, using parameters R = 10.0 and a = 0.5. |
 |
Reconstruction using Hardy's global method, using parameters R = 100.0 and a = 0.5. |
 |
Reconstruction using Hardy's global method, using parameters R = 1.0 and a = 0.1. |
 |
Reconstruction using Hardy's local method, using 10 closest neighbours and parameters R = 1.0 and a = 0.5. |
 |
Reconstruction using Hardy's local method, using 100 closest neighbours and parameters R = 1.0 and a = 0.5. |
 |
Reconstruction using Sibson's method. |
 |
Reconstruction using piecewise linear interpolation based on a Delaunay triangulation of the sample points. |
 |
Reconstruction using piecewise linear interpolation based on an optimized triangulation of the sample points. |