search

MathOnco / valis

Virtual Alignment of pathoLogy Image Series
https://valis.readthedocs.io/en/latest/
MIT License
115 stars 28 forks source link

Issues with large image stack alignment #105

Open pgensbigler opened 6 months ago

pgensbigler commented 6 months ago

We have been hoping to use VALIS to align large image stacks of up to 400 images, but have been running into issues. Specifically, the program will self-terminate. We would appreciate any advice you could lend. Currently, we believe the problem resides in the rigid transformation step as the output mask and early steps appear correct. When we divide the data set into smaller stacks, the program was able to align these smaller subgroups individually but self-terminates when asked to then align the subgroup stacks with each other. We did attempt to reduce the pixels per image in an attempt to smooth the process, but it had no effect.

cdgatenbee commented 6 months ago

Thanks @pgensbigler for creating this issue. Given the large number of images, it sounds like this may be a memory issue. I'll see if I can identify some areas where memory usage can be reduced so that you'll be able to align these larger datasets.

Best, -Chandler

cdgatenbee commented 2 months ago

Hi @pgensbigler, I wanted to let you know the most recent of valis (1.1.0) has (finally) been published. It should address many of the issues you've been running into, including excessive memory usage causing crashes, and poor rigid alignment. Please try it out and let me know how it goes. Thanks again for your patience.

Best,