Image segmentation

This section contains an introdcution into image segmentation, a.k.a. pixel classification as well as creating meshes from segmented data. In particular, this will cover the following topics:

Topic	Description	Link
Segmentation & Surfaces	Introduction to image segmentation and surface extraction & processing	Slides
Thresholding	Basic technique to separate foreground from background	Slides
Instance segmentation	Separating different objects in an image	Slides
Segmentation quality estimation	How to measure the quality of a segmentation	Slides

The course will follow along the slides on Segmentation & Surfaces. The other slides are provided for reference and further explanations, if needed.

Notebooks for exercise

In this session, make sure to execute the following notebooks:

• Thresholding

• Connected components

• Voronoi otsu labelling

• Marching cubes

Further information

Segmentation is a big topic. For further information, also check out the following topics and notebooks:

• Morphological operations

• Otsu thresholding

• Touching objects labelling

• Voronoi tesselation

• Segmentation in 3D

• Segmenting membranes in 2D

• Seeded watershed

• Removing labels at image edges

• Morphological opening and closing

• Segmentation metrics: The segmentation game

• Segmentation metrics: Estimate quality