| |
|
Description: |
Medical image processing remains an active field of research principally due to the difficulty of defining the problem. The most successful approaches are related to solving PDEs that model anisotropic diffusion, where the anisotropy is given spatially by the image gradient or Laplacian, or in direction by identifying principal directions. Other successful methods have used block matching, relating different portions of an image and by comparison estimating the noise and underlying structure. In this presentation the image processing is performed on the image gradient field, which provides a wealth of information regarding the objects and noise in the image. Due to the presence of noise, the image derivatives are obtained using a Least Squares approach or through convolution with a Gaussian function (hence at different scales depending on the standard deviation). The talk is introductory and is intended to attract researchers in different fields to work in image processing.
|
|
Date: |
|
| Start Time: |
14:30 |
|
Speaker: |
Alberto Gambaruto (Barcelona Supercomputer Center, Spain)
|
|
Institution: |
Barcelona Supercomputer Center
|
|
Place: |
Sala 5.5
|
| Research Groups: |
-Analysis
|
|
See more:
|
<Main>
|
|
