Hot to use a trainedProject.lip to generate predictions from within Python, using batch Mode for several input images.jpg present on disk?

There is a"example_python_client.py" that use random input data to “batch process”.

I want to import ilastik on my own ptython project, and generate predictions using a pre trained pixel classification workflow project.lip and several input images.jpg (batch mode I guess…)

Anyone can share sample lines about doing that?

Regards…

While it is possible to do this, we currently do not recommend doing it, as we are in the process of reworking this whole bit, so it will change considerably. So right now I would suggest using ilastik’s headless mode, where you start ilastik from the command line (or via a subprocess in Python).

Hi @ALCTongas

I have a code snippet here that I have created for calling ilastik in headless mode from python

This is still rough around the edges and without documentation but it works for me. Maybe it is useful for you. Currently, only supports pixel classification workflows but other workflows could be added.

1 Like