Name
Abstract | ||
|---|---|---|
We propose a new variational method to isolate points in biological images. As points are fine structures they are difficult to detect by derivative operators computed in the noisy image. In this paper we propose to compute a vector field from the observed intensity so that its divergence explodes at points. As the image could contains spots but also noise and curves where the divergence also blows up, we propose to capture spots by introducing suitable energy whose minimizers are given by the points we want to detect. In order to provide numerical experiments we approximate this energy by means of a sequence of more treatable functionals by a Γ-convergence approach. Results are shown on synthetic and biological images. |
Year | DOI | Venue |
|---|---|---|
2009 | 10.1109/ISBI.2009.5193301 | ISBI |
Keywords | Field | DocType |
new variational method,suitable energy,numerical experiment,convergence approach,fine structure,observed intensity,derivative operator,divergence explodes,biological image,noisy image,probability density function,vector field,noise measurement,biology,data mining,noise,length measurement | Computer vision,Divergence,Noise measurement,Pattern recognition,Variational method,Vector field,Computer science,Length measurement,Γ-convergence,Operator (computer programming),Artificial intelligence,Probability density function | Conference |
Citations | PageRank | References |
0 | 0.34 | 1 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Daniele Graziani | 1 | 0 | 1.35 |
| Laure Blanc-Féraud | 2 | 536 | 63.97 |
| Gilles Aubert | 3 | 1275 | 108.17 |