MegaOCR

MegaOCR is a portable OCR application for Windows powered by Tesseract OCR.
It can read PDF / PNG / JPG and export extracted text to .txt / .docx / .pdf.

👉 Download Latest Portable EXE — no installation required, just run and use.

---

For Users

Features

• Works out of the box on Windows — no Python, no Tesseract installation required.
• Supports 120+ languages (common models already bundled in the Release).
• Exports to multiple formats: .txt, .docx, .pdf.
• Clean and simple interface built with Tkinter (using ttkbootstrap).

Limitations

• OCR accuracy depends on input quality (high-DPI recommended).
• Complex page layouts may reduce accuracy.
• Current PDF export has limited RTL shaping (Persian/Arabic). Use .docx or .txt for best results.

---

For Developers

Run from Source

git clone https://github.com/Megahertz418/MegaOCR.git
cd MegaOCR
pip install -r requirements.txt
python Mega_OCR.py

Make sure Tesseract traineddata files for your target languages are available.

Build (Windows, PyInstaller)

pyinstaller --onefile --noconsole ^
  --add-data "tesseract.exe;." ^
  --add-data "tessdata;tessdata" ^
  --add-data "fonts;fonts" ^
  --add-data "*.dll;." ^
  --add-data "mega_ocr.ico;." ^
  --icon=mega_ocr.ico Mega_OCR.py

The executable will appear at dist/Mega_OCR.exe.

Tip: Use the helper script for clean builds:

scripts\build.ps1 -VendorDir .\vendor

---

Vendor Bundle (for Reproducible Builds)

Each Release ships a Vendor Bundle ZIP, which includes:

• tesseract.exe + required *.dll
• curated tessdata/ models
• fonts/
• mega_ocr.ico and mega_ocr.png
• all third-party licenses
• MANIFEST.json (components + SHA256 hashes)
• SHA256SUMS.txt

📥 Available on the Releases page.

Note: End-users don’t need this ZIP. It is only for developers who want to reproduce the official Release build.

---

Project Layout (source repo)

MegaOCR/
│   .gitignore
│   CHANGELOG.md
│   CONTRIBUTING.md
│   LICENSE
│   Mega_OCR.py
│   Mega_OCR.spec
│   README.md
│   requirements.txt
│   SECURITY.md
│   THIRD_PARTY_NOTICES.md
│
├── .github/               # GitHub-specific configs (PR/Issue templates)
│   │   pull_request_template.md
│   └── ISSUE_TEMPLATE/
│           bug_report.md
│           feature_request.md
│
├── docs/                  # Documentation & media (UI preview, etc.)
│       User Interface.gif
│
└── scripts/               # Build & manifest generation helpers
        build.ps1
        generate-manifest.ps1

Note: vendor/, dist/, and build/ directories are not committed to the repo. They are provided as part of the downloadable Release assets (Vendor Bundle ZIP & EXE).

---

Releases

• Portable EXE (Mega_OCR.exe) — recommended for most users.
• Vendor Bundle ZIP — for developers who want reproducible builds.

Both are available on the Releases page.

---

Troubleshooting

• Empty OCR output: Try higher-resolution images or check language settings (e.g., eng+fas).
• Persian/Arabic shaping in PDF: Use .docx or .txt instead.
• Antivirus false positive: PyInstaller executables are sometimes flagged. Verify integrity with SHA256 checksums in the Vendor Bundle.

Still stuck? Please open an Issue.

---

Roadmap

• macOS/Linux support
• Better complex-layout handling
• Optional CLI mode
• Improved RTL shaping in PDF exports

---

Contributing

We welcome contributions! See CONTRIBUTING.md.

---

Security

If you discover a security issue, please follow the process in SECURITY.md.

---

Changelog

See CHANGELOG.md for version history.

---

Third-Party Notices

See THIRD_PARTY_NOTICES.md for a list of included components. Detailed license texts are provided in the Vendor Bundle ZIP.

---

License

• MegaOCR code: MIT
• Tesseract & models: Apache-2.0 (see THIRD_PARTY_NOTICES.md)
• Fonts: OFL and CC BY 4.0 (see Vendor Bundle)