This is a modified version of Tesseract. The official upstream project is here: https://github.com/tesseract-ocr/tesseract
Dependencies:
- leptonica: (tested with leptonica 1.83+: master branch)
TODO
The modifications in the visible_pdf_image branch enable the user to input both a "cleaned" image to be used for OCR and a "visible" image that is used in the output PDF. Cleaning an image helps OCR engines by removing background colors and patterns, sharpening text, increasing contrast, etc. The process usually makes the image look terrible to humans, so the idea with this fork is to give us the best of both worlds. This is very useful for digitizing documents.
To clean an image for OCR, try using textcleaner from Fred's ImageMagick Scripts: http://www.fmwconcepts.com/imagemagick/textcleaner/
For the visible image, you can use a compressed version to save space. The only requirement is that the dimensions of the "cleaned" and "visible" images are the same.
Once you've built the visible_pdf_image branch along with the other Tesseract dependencies, just add --visible-pdf-image <image> to the arguments. For example:
tesseract -l eng --visible-pdf-image compressed.webp cleaned.pnm out pdf
Here's the original feature request upstream: tesseract-ocr#210
This is a modified version of Tesseract. The official upstream project is here: https://github.com/tesseract-ocr/tesseract
Dependencies:
- leptonica: latest release / master branch.
The modifications in the visible_pdf_image branch enable the user to input both a "cleaned" image to be used for OCR and a "visible" image that the user will see in the output PDF. Cleaning an image helps OCR engines by removing background colors and patterns, sharpening text, increasing contrast, etc. The process usually makes the image look terrible to humans, so the idea with these patches is to give us the best of both worlds.
To clean an image for OCR, I suggest using textcleaner from Fred's ImageMagick Scripts: http://www.fmwconcepts.com/imagemagick/textcleaner/
For the visible image you can use the original copy, but I suggest using a compressed version instead to save space. The only requirement is that the dimensions of the "cleaned" and "visible" images are the same.
Once you've built the visible_pdf_image branch along with the other Tesseract dependencies, just add --visible-pdf-image <image> to the arguments. For example:
tesseract -l eng --visible-pdf-image compressed.webp cleaned.pnm out pdf
I've found this to be very helpful when creating searchable PDFs from scanned documents, receipts, etc. YMMV. Here's the original feature request upstream: tesseract-ocr#210
This package contains an OCR engine - libtesseract and a command line program - tesseract.
Tesseract 4 adds a new neural net (LSTM) based OCR engine which is focused on line recognition, but also still supports the legacy Tesseract OCR engine of Tesseract 3 which works by recognizing character patterns. Compatibility with Tesseract 3 is enabled by using the Legacy OCR Engine mode (--oem 0). It also needs traineddata files which support the legacy engine, for example those from the tessdata repository.
-
Stefan Weil is the current lead developer. Ray Smith was the lead developer until 2018. The maintainer is Zdenko Podobny. For a list of contributors see AUTHORS and GitHub's log of contributors.
-
Tesseract has unicode (UTF-8) support, and can recognize more than 100 languages "out of the box".
-
Tesseract supports various image formats including PNG, JPEG and TIFF.
-
Tesseract supports various output formats: plain text, hOCR (HTML), PDF, invisible-text-only PDF, TSV, ALTO and PAGE.
-
You should note that in many cases, in order to get better OCR results, you'll need to improve the quality of the image you are giving Tesseract.
-
This project does not include a GUI application. If you need one, please see the 3rdParty documentation.
-
Tesseract can be trained to recognize other languages. See Tesseract Training for more information.
Tesseract was originally developed at Hewlett-Packard Laboratories Bristol UK and at Hewlett-Packard Co, Greeley Colorado USA between 1985 and 1994, with some more changes made in 1996 to port to Windows, and some C++izing in 1998. In 2005 Tesseract was open sourced by HP. From 2006 until November 2018 it was developed by Google.
Major version 5 is the current stable version and started with release 5.0.0 on November 30, 2021. Newer minor versions and bugfix versions are available from GitHub.
Latest source code is available from main branch on GitHub. Open issues can be found in issue tracker, and planning documentation.
See Release Notes and Change Log for more details of the releases.
You can either Install Tesseract via pre-built binary package or build it from source.
Before building Tesseract from source, please check that your system has a compiler which is one of the supported compilers.
For Windows user, once you download the Tesseract OCR 4, follow the pop-up instructions, choose which components of the software you would like to install (if storage is limited). Make sure to remember where you saved the file, you will need it to add your environment variables. You will notice that if you try to use tesseract right after the installation it will not be available on the command line (at least not yet). To remedy this, we need to add it the path where the application was saved by taking the following steps:
- Open system environment properties from your settings menu and select environment variable.
- Open file explorer and go the Tesseract folder (should be in C drive by default)
- Once you find the executable (tesserac.exe) fill, copy the path to it.
- Get back to the environment variable window, click path then edit then new path variable and past the path, and save.
- After this, you are good to open a new command line, to make sure it is there, type tesseract, this should open the help page with all the command you can use.
Basic command line usage:
tesseract imagename outputbase [-l lang] [--oem ocrenginemode] [--psm pagesegmode] [configfiles...]
For more information about the various command line options use tesseract --help or man tesseract.
Examples can be found in the documentation.
1- For non-roman alphabet languages (languages with alphabet other abc…) it is best to have the image with the word oriented in a way that makes the words horizontal. This helps prevent errors sometimes. 2- If you are running an image with just one line and it is not recognized, try running again with different psm mode setting, it should output the right words. 3- Do not run an image with no words on it as it will sometimes stall the system.
Developers can use libtesseract C or
C++ API to build their own application. If you need bindings to libtesseract for other programming languages, please see the
wrapper section in the AddOns documentation.
Documentation of Tesseract generated from source code by doxygen can be found on tesseract-ocr.github.io.
Before you submit an issue, please review the guidelines for this repository.
For support, first read the documentation, particularly the FAQ to see if your problem is addressed there. If not, search the Tesseract user forum, the Tesseract developer forum and past issues, and if you still can't find what you need, ask for support in the mailing-lists.
- tesseract-ocr - For tesseract users.
- tesseract-dev - For tesseract developers.
Please report an issue only for a bug, not for asking questions.
The code in this repository is licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
NOTE: This software depends on other packages that may be licensed under different open source licenses.
Tesseract uses Leptonica library which essentially uses a BSD 2-clause license.
Tesseract uses Leptonica library for opening input images (hence not documents like pdf). It is suggested to use leptonica with built-in support for zlib, png and tiff (for multipage tiff).
For the latest online version of the README.md see:
https://github.com/tesseract-ocr/tesseract/blob/main/README.md