maxschelski / livecellreg

Correlation-based registration of 2D & 3D live cell imaging movies in x, y and z axis.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

livecellreg

Correlation-based registration of 2D & 3D live cell imaging movies using translation of images. It also works quite well for cells with signals that change localization over time within the structure that should be registered - e.g. a signal that is moving within the cell. Only works with timelapse recording - input tif files must have multiple frames as set in ImageJ under Image -> Properties.

Still needs a lot of refactoring. Will be converted into a locally installable package.

For now do the following to use the script

  1. Check parameters outlined and described at the beginning of the script. If results are not good think about changing parameters - e.g. the frequency of getting a new reference image or for large shifts the max_shift value.
  2. execute script livecellreg.py with
  • choose_folder_manually=True and
  • for multi channel movies set the channel on which the registration should be based as reference_channel = "c000X" where X is the channel number (counting starts at 0).
  1. A popup window will open in which you can choose in which folder you want to register all tiff files.

For questions, please write to me at max.schelski@googlemail.com.

About

Correlation-based registration of 2D & 3D live cell imaging movies in x, y and z axis.

License:GNU General Public License v3.0


Languages

Language:Python 100.0%