organoid_tracker.py cannot recognize .tif, .jpg, .gif files

[hanmingcr]

The package was installed on Ubuntu 20.04. Unfortunately, the executable file cannot recognize any regular image files.

[rutgerkok]

That's strange. The program should start just by running python organoid_tracker.py, and then the screen below should open. From there, you can load a lot of different image formats.

That said, we have never tested the program on Ubuntu, only on Windows. So I expect some issues to arise. If you encounter any errors in the console, you can post them here and I'll have a look.

[hanmingcr]

Thank you for your reply. There is no error message when I run organoid_tracker.py. It just cannot recognize any image, even the ones included in the manuals/images/ folder (see the attached screenshot).

[rutgerkok]

Thanks for the screenshot! Looking at the official documentation ( https://doc.qt.io/qtforpython-5/PySide2/QtWidgets/QFileDialog.html#PySide2.QtWidgets.PySide2.QtWidgets.QFileDialog.getOpenFileName ) it seems like I need to use a space instead of a semicolon to separate the file extensions (*.nd2;*.tif). Maybe that fixes it? I've updated OrganoidTracker to include this change.

[hanmingcr]

Thank you! The issue is gone now. By the way, would you mind uploading a python script to convert the checkpoint files of your pre-trained model into saved_model.pb? I used the savedmodel module to convert, but it generates a .pb file incompatible to tensorflow2.5.

[rutgerkok]

I would use the stable branch - https://github.com/jvzonlab/OrganoidTracker/tree/stable . That one contains the network as it was published.

For the master branch, we're still testing things out, to find what works best.