Multiresolution segmentation of natural images: from linear to nonlinear scale-space representations.
- Petrovic A Signal Processing Institute, Swiss Federal Institute of Technology, Lausanne, Switzerland. ana.petrovic@epfl.ch
- Escoda OD
- Vandergheynst P
- 2004-08-26
Published in:
- IEEE transactions on image processing : a publication of the IEEE Signal Processing Society. - 2004
Algorithms
Image Enhancement
Image Interpretation, Computer-Assisted
Linear Models
Nonlinear Dynamics
Pattern Recognition, Automated
Reproducibility of Results
Sensitivity and Specificity
Signal Processing, Computer-Assisted
English
In this paper, we introduce a framework that merges classical ideas borrowed from scale-space and multiresolution segmentation with nonlinear partial differential equations. A non-linear scale-space stack is constructed by means of an appropriate diffusion equation. This stack is analyzed and a tree of coherent segments is constructed based on relationships between different scale layers. Pruning this tree proves to be a very efficient tool for unsupervised segmentation of different classes of images (e.g., natural, medical, etc.). This technique is light on the computational point of view and can be extended to nonscalar data in a straightforward manner.
- Language
-
- English
- Open access status
- green
- Identifiers
-
- DOI 10.1109/tip.2004.828431
- PMID 15326852
- Persistent URL
- https://sonar.ch/global/documents/77995
Statistics
Document views: 26
File downloads:
- Full-text: 0