Skip to content

OCR

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

OCR

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
README.adoc
README.adoc
 
 
ocr_gui.sh
ocr_gui.sh
 
 
setup_ocr.sh
setup_ocr.sh
 
 

README.adoc

OCRmyPDF + Tesseract — Installation & GUI Wrapper

Table of Contents

This folder contains two complementary scripts:

  • setup_ocr.sh – Installation wizard for OCRmyPDF, Tesseract OCR and language packs.

  • ocr_gui.sh – Graphical (Zenity) frontend to OCR scanned PDFs with full navigation, settings, and batch processing.

They are designed to work as a single unit: run the setup wizard once, then use the GUI for daily OCR work.

1. Setup Wizard (setup_ocr.sh)

Installs OCRmyPDF and Tesseract OCR on Fedora or Ubuntu/Debian.

1.1. Features

  • Distribution Detection:: Automatically detects Fedora or Ubuntu/Debian and uses the correct package manager (dnf / apt).

  • Interactive Wizard:: Three-step guided setup: base install, language-pack selection, usage examples.

  • State Tracking:: On re-runs, skips the base installation and jumps straight to language-pack selection.

  • Health Check:: Detects missing components (e.g., after manual uninstall) and repairs them automatically.

  • 30 Language Packs:: Supports English, German, French, Spanish, Russian, Chinese, Japanese, and many more.

  • --full Flag:: Force re-run the complete wizard including base installation.

1.2. Usage

sudo ./setup_ocr.sh
sudo ./setup_ocr.sh --full    # force full reinstall

1.3. What It Installs

Package

Fedora

Ubuntu

ocrmypdf

ocrmypdf

ocrmypdf

Tesseract

tesseract

tesseract-ocr

Language packs

tesseract-langpack-*

tesseract-ocr-*

Orientation & Script Detection

tesseract-osd

tesseract-ocr-osd

GUI dialogs

zenity

zenity

Scan cleanup

unpaper

unpaper

1.4. Troubleshooting

  • Missing osd.traineddata → run sudo ./setup_ocr.sh (health check will repair)

  • Ubuntu — universe repository required → sudo add-apt-repository universe

  • OCR quality is poor → install additional language packs or adjust GUI settings

2. Graphical OCR Frontend (ocr_gui.sh)

A Zenity-based GUI wrapper around ocrmypdf that provides a complete step-by-step workflow.

2.1. Workflow

Mode selection → File/Folder picker → Language selection
→ Text mode → Naming mode → OCR processing → Result dialog

Every step has a Back button; cancel on the main screen exits.

2.2. Features

  • Single File & Batch Mode:: Process one PDF or a whole folder at once.

  • Language Selection:: Dynamically reads installed Tesseract languages (dynamically detected at startup).

  • Text Mode:: --skip-text (faster) or --force-ocr (rebuild text layer) with optional "Remember" setting.

  • Output Naming:: Three options — replace original, add _OCR suffix, or auto-indexed numbers (_1, _2, …).

  • Remembered Settings:: Text mode and naming choices can be saved to ocr_settings.json (lives next to the script, manageable via Settings window).

  • Batch Processing:: Processes all unprocessed PDFs in a folder with a progress bar.

  • Error Logging:: Timestamped debug output visible directly in the terminal.

  • Navigation:: "OCR another file" restarts the workflow at the file picker.

2.3. Usage

cd Program_Setups/OCR
./ocr_gui.sh

Optional: enable full bash trace for debugging:

OCR_DEBUG=1 ./ocr_gui.sh

2.4. Dependencies

Installed by setup_ocr.sh:

  • zenity – GUI dialogs

  • ocrmypdf – OCR engine

  • tesseract – OCR backend

  • tesseract-osd – Orientation & Script Detection

  • unpaper – Scan image cleanup (for --clean flag)

  • python3 – JSON settings management

  • bash ≥ 4.0 – associative arrays and mapfile

2.5. Settings Window

Accessed from the main menu. Shows all configurable options:

  • Text handling mode – --skip-text or --force-ocr (checkbox, rememberable)

  • Naming: Replace original file – overwrites the source PDF

  • Naming: _OCR suffix – appends _OCR to the filename (default)

  • Naming: Index number – creates filename_1.pdf, filename_2.pdf, …

Naming modes are mutually exclusive: only the last checked one is kept on Save.

2.6. Example Session

$ ./ocr_gui.sh
[10:15:00] DIALOG: mode selection
→ Pick "Process single PDF file"
→ Select a PDF
→ Pick language "deu"
→ Text mode: --skip-text
→ Naming: _OCR suffix
→ OCR runs… success!
→ "OCR another file?" → back to file picker

3. Files

File

Purpose

setup_ocr.sh

Installation wizard (must be run as root)

ocr_gui.sh

Graphical OCR frontend

ocr_settings.json

Remembered settings (auto-created)

README.adoc

This file