AmhOCR

Tesseract Powered Windows Desktop OCR Application With Multiple Pre/Post Processing GUI

AmhOCR is an Optical Character Recognition (OCR) application for Windows Desktop. It is based on the latest Tesseract's v4.0/LSTM OCR engine which supports over 100 languages. However some post-processing tools, in AmhOCR, are applicable only for Amharic language.

Test Screenshot: Multi column newspaper image recognition, with image area embedding, spell checking(highlighted in red) and word level text editing.

Features:

• Support OCRing multiple image format: TIFF, JPEG, PNG, BMP and other image formats
• Import PDF as image
• Batch image import
• Flexible image view tool
• Pre-process image: de-skew, threshold and other image correction settings
• Process selected image
• Batch Process image with flexible management: pause, re-start and cancel
• Asynchronous and multithread processing for large batch size
• View and work on the processed text with similar format and layout as source input
• Post Process OCR output: Edit text and spell check
• Multi level editing: Paragraph level and word level edit.
• Convert OCR output: to MS word document, searchable PDF and text output
• Save and Open work as AmhOCR single file project, based on HOCR data structure
• NLP process: word list, word frequency, sentences list, word & character N-gram
• Extension tools for multipage TIFF: Merge and split
• Extension tools for PDF: Merge, split and convert to image
• Extension tools for image: Format conversion to different image format and conversion to PDF

Dependencies:

• Microsoft .NET Framework 4.0
• GhostScriptSharp
• AForge Imaging
• iTextSharp
• Microsoft.Bcl.Async
• OpenXML SDK 2.5

Building with Visual Studio:

• Visual studio 2015 and later
• Restore nuget packages

License:

• GNU GPLv3