Image annotations on large file


I am wondering whether someone has experience in annotating large image files. I have some z-stacked CZI files from the Zeiss AxioScan Z1 and I want to do pixel annotations of my objects for training a mask-c-rnn network.
Does anyone know a nice annotation tool which could handle large size images? It’s not necessary to handy CZI files, I could convert the image into standard image file formats.

Any help would be appreciated

Not sure whether you are aware of napari. As your post is tagged “python” that would be the first suggestion.

It is also not quite clear whether you need to do the annotations in 3D or individual slices.
For very large volumes there might be some lag. I believe this is currently being worked on by @DragaDoncila and @jni .


Depending on the size, QuPath might be an option, as it is designed for “large” files. Though I tend to think of large for QuPath being a few gigabytes, as opposed to a terabyte for lightsheet data. It really depends on what you want to do.

A popular thread of some suggestions for annotating 3D data sets:


Thanks I have heard of Napari, I am going to try it out. I need only the positions of the labels of one slice, since I can use the same coordinates for all slides.

Thanks, I have about 10-50 GB files. I will read the thread :+1: