Pixel classifier traing on multiple images

Hi, I am exploring QuPath Pixel classifier and I wanted to train my classifier on roi from different images. The problem is that when I save it I cannot reprise resume training and I can only apply it. Is there a way to train on different images?

Create annotations for the areas you want to pull from using the same class for each of these annotations. Usually it would be the Region* class.
Note that if the image areas are too large, this could cause your computer some problems. Depends on the memory available, I think.
If you have many images, you can make the sub regions wrap around by differently by changing the pixel width.

Thank for the reply.
Yes, I solved with “Region” annotations

1 Like