Name
Abstract | ||
|---|---|---|
Many nano-scale sensing techniques and image processing applications are characterized by noisy, or corrupted, im- age data. Unlike typical camera-based computer vision im- agery where noise can be modeled quite well as additive, zero-mean white or Gaussian noise, nano-scale images suf- fer from low intensities and thus mainly from Poisson-like noise. In addition, noise distributions can not be consid- ered symmetric due to the limited gray value range of sen- sors and resulting truncation of over- and underflows. In this paper we adapt B-spline channel smoothing to meet the requirements imposed by this noise characteristics. Like PDE-based diffusion schemes it has a close connection to robust statistics but, unlike diffusion schemes, it can handle non-zero-mean noises. In order to account for the multi- plicative nature of Poisson noise the variance of the smooth- ing kernels applied to each channel is properly adapted. We demonstrate the properties of this technique on noisy nano-scale images of silicon structures and compare to anisotropic diffusion schemes that were specially adapted to this data. |
Year | DOI | Venue |
|---|---|---|
2003 | 10.1109/CVPRW.2003.10018 | CVPR Workshops |
Field | DocType | Volume |
Value noise,Computer vision,Colors of noise,Noise measurement,Computer science,Salt-and-pepper noise,Image noise,Smoothing,Artificial intelligence,Gaussian noise,Gradient noise | Conference | 2 |
Issue | ISSN | Citations |
1 | 1063-6919 | 3 |
PageRank | References | Authors |
0.52 | 11 | 3 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Hanno Scharr | 1 | 430 | 37.92 |
| Michael Felsberg | 2 | 2419 | 130.29 |
| Per-Erik Forssen | 3 | 63 | 4.41 |