Trainable segmentation

Your support by citing our work will motivate us to release more free code:

Burget, R., Karásek, J., Smékal, Z., Uher V., Dostál, O., RapidMiner Image Processing Extension:A Platform for Collaborative Research, International Conference  on TELECOMMUNICATIONS AND SIGNAL PROCESSING, Baden Austria 2010

Description

This page describes a method for trainable segmentation. According to a given image and given segmented result, the algorithms described here can be used for trainable image segmentation. The trained result can be applied to other images, where the segmentation will be similar to the trained method. The paper related to this work was related to segmentation of functional parts of animal brain, however I hope it can be applied also to many different areas. These algorithm was used on the ISBI challenge 2012, (http://brainiac.mit.edu/isbi_challenge/).

Screenshots

 Download

RapidMiner Book

Introduction to RapidMiner & IMMI

This tutorial is really a brief and basic introduction into RapidMiner and IMMI and it demonstrates how to load images using RapidMiner and IMMI extension.

You need to install or upgrade Flash Player to view this content, install or upgrade by clicking here.

Chapter 3.1 Point selection

This tutorial demonstrates how to select points for training. There are three basic ways: regular grid, random or using point of interest algorithms. A brief introduction how to select points with RapidMiner is provided by this tutorial.

You need to install or upgrade Flash Player to view this content, install or upgrade by clicking here.

Chapter 3.2 Local-feature extraction & Chapter 3.3 Trainable segmentation

This tutorial demonstrates how to use “trainable segmentation” operator. It demonstrates 1) how to create different different transform, 2) how to use different learning algorithms.

You need to install or upgrade Flash Player to view this content, install or upgrade by clicking here.