|
Abstract
|
|
|---|---|
| Abstract. The segmentation of medical images poses a great challenge in the area of image processing and analysis due mainly to noise, com- plex background, fuzzy and overlapping objects, and non-homogeneous gradients. This work uses the so-called locally constrained watershed transform introduced by Beare [1] to address these problems. The shape constraints introduced by this type of flexible watershed transformation permit to successfully segment and separate regions of interest. This type of watershed offers an alternative to other methods (such as distance function flooding) for particle extraction in medical imaging segmenta- tion applications, where particle overlapping is quite common. Cytology images have been used for the experimental results. | |
|
International
|
Si |
|
Congress
|
10th symposium on Mathematical Morphology |
|
|
960 |
|
Place
|
Intra, Lago Maggiore, Italia |
|
Reviewers
|
Si |
|
ISBN/ISSN
|
978-3-642-21568-1 |
|
|
10.1007/978-3-642-21569-8 |
|
Start Date
|
06/07/2011 |
|
End Date
|
08/07/2011 |
|
From page
|
429 |
|
To page
|
438 |
|
|
Mathematical Morphology and Its Applications to Image and Signal Processing. 10th International Symposium, ISMM 2011 Verbania-Intra, Italy, July 6-8, 2011 Proceedings Eds.: Pierre Soille, Martino Pesaresi, Georgios K. Ouzounis |