Use better tesseract training dataset #459
Merged
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
This was referenced Dec 5, 2024
jonchang
force-pushed
the
use-better-tesseract-data
branch
2 times, most recently
from
599cca6 to
90e730c
Compare
This should improve OCR performance using tesseract with only a mild increase in runtime and container size.
jonchang
force-pushed
the
use-better-tesseract-data
branch
from
164cb20 to
b888f33
Compare
arinkulshi-skylight
approved these changes
Dec 9, 2024
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Outdated suggestions cannot be applied.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.
Description
Uses a better tesseract training dataset.
By default, Debian (and Ubuntu) use the
tesseract_fasttraining data for their repositories. https://github.com/AlexanderP/tesseract-lang-debian/blob/HEAD/debian/upstream/metadataThis pull request downloads the
tesseract_bestdataset, which is slightly larger and has a roughly 2x performance penalty, but in practice this isn't a big deal (5 mb -> 12 mb, 0.75 s -> 1.5 s).Also update to tesseract v5 and make some changes to slim down the docker image.
Related Issues
Part of #422. This was used to generate benchmark metrics for comparisons locally, so this pull request will port those changes to the deployed image.
Also part of #412. Moving to python-slim (and cleaning up poetry caches) had the effect of dropping the base image size from 3gb to 2gb.
Checklist