Binarize with Olena/scribo
make deps-ubuntu
...will try to install the required packages on Ubuntu.
make build-olena
...will download, patch and build Olena/scribo from source,
and install its standalone CLI scribo-cli (see below)
locally (in $VIRTUAL_ENV or in $PREFIX if given).
make install
...will run build-olena, if necessary, and install the Python
package ocrd_olena with the OCR-D CLI
ocrd-binarize-olena (see below).
make test
...will clone the assets repository from Github, make a workspace copy, and run checksum tests for binarization on them.
This package has the following user interfaces:
Converts images in any format.
Usage: scribo-cli COMMAND [ARGS]
List of available COMMAND options:
Full Toolchains
---------------
* On documents
doc-ppc Common preprocessing before looking for text.
doc-ocr Find and recognize text. Output: the actual text
and its location.
doc-dia Analyse the document structure and extract the
text. Output: an XML file with region and text
information.
* On pictures
pic-loc Try to localize text if there's any.
pic-ocr Localize and try to recognize text.
Tools
-----
* xml2doc Convert the XML results of document toolchains
into user documents (HTML, PDF...).
Algorithms
----------
* Binarization
otsu Otsu's (1979) global thresholding algorithm.
niblack Niblack's (1985) local thresholding algorithm.
sauvola Sauvola and Pietikainen's (2000) local/adpative algorithm.
kim Kim's (2004) algorithm.
wolf Wolf and Jolion's (2004) algorithm.
sauvola-ms Lazzara's (2013) multi-scale Sauvola algorithm.
sauvola-ms-fg Extract foreground objects and run multi-scale
Sauvola's algorithm.
sauvola-ms-split Run multi-scale Sauvola's algorithm on each color
component and merge results.
singh Singh's (2014) algorithm.
Other
-----
version Show version and exit
help Show this message and exit
For command arguments, see 'scribo-cli COMMAND --help' for more information
on each specific COMMAND.
For example:
scribo-cli sauvola-ms path/to/input.tif path/to/output.png --enable-negate-output
This can also be used with the general-purpose image preprocessing OCR-D wrapper ocrd-preprocess-image to get the power of Olena's binarization to all structural levels of the PAGE segment hierarchy. (See this parameter preset for an usage example.)
OCR-D processor interface ocrd-olena-binarize
To be used with PageXML documents in
an OCR-D annotation workflow. Input could be any valid workspace
with source images available. Covers PAGE hierarchy levels page, table, region and
line.
Uses either (the last) AlternativeImage/@filename (if any), or Page/@imageFilename
(otherwise, cropping to Border if necessary). Adds an AlternativeImage with the
result of binarization for every segment.
Usage: ocrd-olena-binarize [worker|server] [OPTIONS]
popular binarization algorithms implemented by Olena/SCRIBO, wrapped for OCR-D
> binarization with Scribo from Olena suite
> For each page, open and deserialize PAGE input file (from existing
> PAGE file in the input fileGrp, or generated from image file).
> Retrieve its respective image at the requested `level-of-operation`
> (ignoring annotation that already added `binarized`).
> Passes the image file to the Olena suite's scribo binarization
> program for the selected algorithm `impl` and its parameters.
> If binarization returns with a failure, skip that segment with an
> approriate error message. Otherwise, put the resulting PNG image
> file into the output fileGrp, and reference it in the METS using a
> file ID with suffix ``.IMG-BIN``. Reference it as AlternativeImage
> in the page, adding ``binarized`` to its @comments.
> Produce a new PAGE output file by serialising the resulting
> hierarchy.
Subcommands:
worker Start a processing worker rather than do local processing
server Start a processor server rather than do local processing
Options for processing:
-m, --mets URL-PATH URL or file path of METS to process [./mets.xml]
-w, --working-dir PATH Working directory of local workspace [dirname(URL-PATH)]
-I, --input-file-grp USE File group(s) used as input
-O, --output-file-grp USE File group(s) used as output
-g, --page-id ID Physical page ID(s) to process instead of full document []
--overwrite Remove existing output pages/images
(with "--page-id", remove only those).
Short-hand for OCRD_EXISTING_OUTPUT=OVERWRITE
--debug Abort on any errors with full stack trace.
Short-hand for OCRD_MISSING_OUTPUT=ABORT
--profile Enable profiling
--profile-file PROF-PATH Write cProfile stats to PROF-PATH. Implies "--profile"
-p, --parameter JSON-PATH Parameters, either verbatim JSON string
or JSON file path
-P, --param-override KEY VAL Override a single JSON object key-value pair,
taking precedence over --parameter
-U, --mets-server-url URL URL of a METS Server for parallel incremental access to METS
If URL starts with http:// start an HTTP server there,
otherwise URL is a path to an on-demand-created unix socket
-l, --log-level [OFF|ERROR|WARN|INFO|DEBUG|TRACE]
Override log level globally [INFO]
--log-filename LOG-PATH File to redirect stderr logging to (overriding ocrd_logging.conf).
Options for information:
-C, --show-resource RESNAME Dump the content of processor resource RESNAME
-L, --list-resources List names of processor resources
-J, --dump-json Dump tool description as JSON
-D, --dump-module-dir Show the 'module' resource location path for this processor
-h, --help Show this message
-V, --version Show version
Parameters:
"level-of-operation" [string - "page"]
PAGE XML segment hierarchy level to annotate images for
Possible values: ["page", "table", "region", "line"]
"impl" [string - "sauvola-ms-split"]
The name of the actual binarization algorithm
Possible values: ["sauvola", "sauvola-ms", "sauvola-ms-fg", "sauvola-
ms-split", "kim", "wolf", "niblack", "singh", "otsu"]
"k" [number - 0.34]
Sauvola's formulae parameter (foreground weight decreases with k);
for Multiscale, multiplied to yield default 0.2/0.3/0.5; for Singh,
multiplied to yield default 0.06; for Niblack, multiplied to yield
default -0.2; for Wolf/Kim, used directly; for Otsu, does not apply
"win-size" [number - 0]
The (odd) window size in pixels; when zero (default), set to DPI (or
301); for Otsu, does not apply
"dpi" [number - 0]
pixel density in dots per inch (overrides any meta-data in the
images); disabled when zero
Copyright 2018-2023 Project OCR-D
ocrd_olena is released under the GNU General Public Licence. See the file
LICENSE (at the root of the source tree) for details.