wip: trying out bbdt for 2D boolean case #42
Draft
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The last version was broken and slow
william-silversmith
force-pushed
the
master
branch
from
cfe3d3e
to
07d5f18
Compare
william-silversmith
force-pushed
the
master
branch
from
3a4cc4e
to
4648385
Compare
Only write one label for every 4 in the first pass.
william-silversmith
force-pushed
the
master
branch
from
40c3c29
to
9943790
Compare
william-silversmith
force-pushed
the
master
branch
2 times, most recently
from
2292fec
to
d8d56df
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Try using BBDT for the 2D boolean situation. This seems to work well for medical images in YACCLAB but on more noisy images, there is no change or a decrease in performance. I'll have to study this more.
For reference, this implementation of BBDT (the 2009 version) clocks about 320 MVx/sec vs the Optimized BBDT OpenCV version which is >500 MVx/sec on medical images. Previously, we were doing 280 MVx/sec (1.14x improvement).
I'm interested in seeing if this technique can be profitably extended to trinary images as well.