TimageTK User Guide

This guide introduce the most important features of TimageTK and is intended as an introductory overview.

For detailed examples, see: Examples.

This is the TimageTK User Guide for TimageTK version .

Components package

The Components module contains both image data structures (SpatialImage, LabelledImage, TissueImage) and functionalities dedicated to input/output operations (imread, imsave).

It consists in four modules, one dedicated to Input/Ouput (see Input/Output module), and three classes:

  • SpatialImage data structure is dedicated to grayscale intensity images (see SpatialImage class)
  • LabelledImage data structure is dedicated to segmented images (see LabelledImage class)
  • TissueImage data structure is also dedicated to segmented images, but with a biology oriented terminology (see TissueImage class)

Input/Output module

The Input/Output module contains functions used to read and write images and transformations. Briefly, use imread or imsave with images and read_trsf or save_trsf with transformations.

|POSS_EXT| replace:: :io.io_image.POSS_EXT


Accepted file formats are: |POSS_EXT|.

Algorithms module

The Algorithms module contains all algorithms wrapped by the plugins (see Plugins module) and more.


These are low level functions, offering a lot of control on parameters, we advise non-expert users to first work with the plugins module.

Plugins module

The Plugins module contains a set of plugins whose API (Application Programming Interface) is standardized. They implement higher level functionalities and provide an access to different methods.

The list of available plugins with their number of implemented methods is:

Linear Filtering:
 9 methods, see Linear Filtering
Grayscale Morphology:
 12 methods, see Grayscale Morphology
Averaging:5 methods, see Averaging
H-transform:2 methods, see H-transform
Region Labeling:
 1 method, see Region Labeling
Segmentation:1 method, see Segmentation
Labels Post Processing:
 ?? methods, see Labels Post Processing
Registration:3 methods, see Registration
Sequence Registration:
 3 methods, see Sequence registration
MultiAngle Fusion:
 1 method, see Fusion


We use the ctypes library to map C-object structure to Python classes. See the Wrapping of Morpheme C library to understand how it was done.