Machine learning to segment neutron images
| Status | Downloads |
|---|---|
 |
   |
Getting started
If you want to run the notebook on your own computer, you’ll need to perform the following step:
- You will need to install Anaconda
- Clone the lecture repository (in the location you’d like to have it)
git clone https://github.com/ImagingLectures/MLSegmentation4NI.git - Enter the folder ‘MLSegmentation’
- Create an environment for the notebook
conda env create -f environment.yml -n MLSeg4NI - Enter the environment
conda activate MLSeg4NI -
Start jupyter and open the notebook
lecture/ML4NeutronImageSegmentation.ipynb -
Use the notebook
- Leave the environment
conda deactivate
Lecture outline
Introduction
- Introduction to neutron imaging
- Some words about the method
- Contrasts
- Introduction to segmentation
- What is segmentation
- Noise and SNR
- Problematic segmentation tasks
- Intro
- Segmenation problems in neutron imaging
Limited data problem
- Training data from NI is limited
- Augmentation
- Transfer learning
Unsupervised segmentation
- e.g. k-means
Supervised segmentation
- e.g. k-NN, decision trees
- NNs for segmentation
Final problem: Segmenting root networks in the rhizosphere using convolutional NNs
- Problem definition
- NN model
- Loss functions
- Training
- Results