bioimaging and optics platform BIOP

TOP Image Splash

Color Segmentation

Student: Joël Leuenberger,

This ImageJ plugin allows for color image segmentation based on user-defined input. The user samples the colors he wants to separate and the plugin provides a segmented view.

Description

In the field of classical histological staining, it is often necessary to characterize different kinds of tissue in the acquired transmission images.

"Color Segmentation" is a plugin developed for color-based images pixels classification, for the public-domain image analysis package ImageJ.
The software implements two different algorithms for pixel clustering based on the distribution of the pixels in the color space and some spatial constraint.
The cluster can be initialized ad hoc by the user in order to get better segmentation results.

Some Image

Example of segmented dataset using both algorithms: The Hidden Markov Model and the K-Means.


Downloads

Download Documentation (PDF)

In collaboration with BioMedical Imaging Group 2006-2008, Dr. Daniel Sage, Sarria JC Floyd, Alessandra Griffa

Figures

Another Figure

Fig. 1: Another output. The Plugin provides the binary masks of the different classes

References

No References as of yet



Contact

BioImaging and Optics platform (PTBIOP)

EPFL-SV-PTECH-BIOP
Station 15
CH-1015 Lausanne
Show on campus map


Dr. Arne Seitz

Office AI-0240
Tel: +41 (0) 21 693 96 18
Fax: +41 (0) 21 693 95 85