Skip to content

Conversation

@parabolala
Copy link
Contributor

This PR includes #2 rebased on top of current HEAD + a commit that cleans up the code and adds an improved adaptive thresholding routine.

Copy link
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Will it help to tune it if I provide you some manually classified shreds? If yes — please let me know what kind of info do you need.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not sure how classified data would help here. Will think about it.

Copy link
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If we can count f1 score for each combination of input params on a known dataset then we can adjust params to achieve the best score possible.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not sure about effort/benefit of this. Might make sense to just leave these params loose and add additional posterior filtering.

@parabolala
Copy link
Contributor Author

Now compatible with master.

@parabolala
Copy link
Contributor Author

PTAL. Line angle estimation added.

dchaplinsky added a commit that referenced this pull request Nov 2, 2014
@dchaplinsky dchaplinsky merged commit 38a8178 into dchaplinsky:master Nov 2, 2014
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.

3 participants