Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

WIP add normalize op #136

Closed
wants to merge 1 commit into from
Closed

WIP add normalize op #136

wants to merge 1 commit into from

Conversation

hinerm
Copy link
Member

@hinerm hinerm commented Mar 14, 2024

No description provided.

@hinerm hinerm marked this pull request as draft March 14, 2024 18:55
@hinerm
Copy link
Member Author

hinerm commented Mar 14, 2024

I started this branch to work on scijava/scijava#197

However, this implementation of normalization is not really what I wanted. In my sample data, the 16-bit histogram looks like this:
image

Really what I want is to normalize in this actual min-max range. Normalizing in the input (16-bit) range to the output (8-bit range results in all of these values being mapped essentially to 0-4ish which is.. not great.

I don't think using a single type-based op and lifting will be appropriate because we need to know the range. We would need a new Op that operates on images and has a dependency on min/max ops and normalize ops.

Normalize would also need to be updated to add signatures that accept input range parameters

@hinerm hinerm force-pushed the normalize-clip-cast branch from a451a10 to 25383e9 Compare March 26, 2024 16:29
@hinerm
Copy link
Member Author

hinerm commented Apr 1, 2024

Found the image.normalize ops and they appear to work for my use case so I am going to just use those instead

@hinerm hinerm closed this Apr 1, 2024
@hinerm hinerm deleted the normalize-clip-cast branch May 2, 2024 18:49
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant