Tesseract 4.0 + source code is available in the ‘master’ branch of the repository. It adds a new OCR engine based on LSTM neural networks. It initially works (well) on x86/Linux. Model data for 101 languages is available in tessdata, tessdata_best, tessdata_fast repositories.
- VGSLSpecs info from Tensorflow
- 4.0 Accuracy and Performance
Training Tesseract LSTM engine
Ubuntu PPAs for Tesseract 4.x & Leptonica 1.7x:
Leptonica 1.74.1 package for Debian:
4.x for Windows
Unofficial experimental binaries of tesseract-ocr 4.x are available from the following links. Each one is from a different commit from master branch in early 2017. See individual sites for more details:
- Windows Installer made with MinGW-w64 from UB Mannheim
- zip file with cppan generated .dll and .exe files, You have to install VC2015 x86 redist from microsoft.com in order to run them.
- Win64 build of tesseract 4.0.0 alpha, leptonica 1.74.1, and charlesw/tesseract .Net wrapper - built using CPPAN for Visual Studio 2017.
4.x with GUI frontend
Windows binaries of tesseract-ocr 4.0.0-alpha with GUI interface are available for VietOCR from
Visual C++ Redistributable for Visual Studio 2015 runtime - vc_redist.x86.exe is REQUIRED for VietOCR to run correctly.
VietOCR can be used to download appropriate 4.0.0alpha traineddata for additional languages.
Windows binaries of tesseract-ocr 4.0.0-alpha with GUI interface are available for gImageReader from
Download 4.0.0alpha traineddata to use with the above from master branch of tessdata. e.g. for Hindi download the following file:
The [3.05 branch on GitHub] (https://github.com/tesseract-ocr/tesseract/tree/3.05) can be used by those who want the bug fixes for 3.05.01 release.
An installer for Tesseract 3.05 for Windows is available from Tesseract at UB Mannheim. This includes the training tools.
Current official release
The current official release is 4.1.1.