A plugin for the DMX platform. It makes the Tesseract OCR engine available on the DMX platform.
The DMX Tesseract plugin relies on a local native Tesseract installation.
Released versions are available at:
https://download.dmx.systems/plugins/dmx-tesseract/
Snapshot versions (work in progress) are available at:
https://download.dmx.systems/ci/dmx-tesseract/
As with any DMX plugin you install it by copying the respective .jar file to DMX's bundle-deploy/ directory.
| Property | Required | Description |
|---|---|---|
| dmx.tesseract.data_path | yes | Path to the tessdata directory of your local Tesseract installation, e.g. /usr/share/tesseract-ocr/tessdata |
After configuration a restart of the DMX platform is required.
There is neither a GUI nor a HTTP interface. Instead the DMX Tesseract plugin provides an OSGi service (systems.dmx.tesseract.TesseractService) to be consumable by other DMX plugins.
The service consists of only one method:
String doOCR(String repoPath)
The image or PDF file to be OCRed is expected to exist in the DMX file repository. The result is the recognized text.
For accessing the native Tesseract (C++) API from Java, Tess4j is utilized, a JNA based wrapper.
The DMX Tesseract service is consumed e.g. by the dmx-pdf-search plugin.
1.0 -- Dec 18, 2023
- Provides a service with a
doOCR()method. - Compatible with DMX 5.3.4 or later