Releases: OCR-D/ocrd_typegroups_classifier
Releases · OCR-D/ocrd_typegroups_classifier
v0.5.0
Added:
- New line-based classification, #14
- Resources now listed in the
ocrd-tool.json
, #15
v0.4.0
Added:
- Support for classifying of sub-page elements, like regions, directed by parameter
level
, #12
- List of fonts to recognize can be restricted with new
active_classes
parameter, #12
Fixed:
typegroups-classifier
simple CLI working again, #11
v0.3.0
Changed:
- Use
self.resolve_resource
to resolve network
parameter, #8
v0.2.0
Changed:
- default model is now bundled with the processor, #7
v0.1.4
Fixed:
- imports in simple CLI, #5
--help
/-h
for simple CLI, #6
v0.1.3
Fixed:
- pass on pageID, ht @mikegerber
- use core's image API (
image_from_page
), ht @bertsky
- annotate
pcGtsId
, Metadata
, document processor, ht @bertsky
v0.1.2
Fixed:
getLogger
mustn't be called in constructor, #3
v0.1.1
Fixed:
- missing
import json
from setup.py
v0.1.0
Changed:
- merge upstream, #2
- use
make_file_id
and assert_file_grp_cardinality
, #2