tesseract-ocr/tessdoc

Tesseract documentation

What it solves

Tesseract is an open-source engine that performs Optical Character Recognition (OCR), allowing computers to extract printed text from images. It provides the necessary tools for developers and users to convert scanned documents or image files into machine-readable text across a wide variety of languages.

How it works

The engine uses an LSTM (Long Short-Term Memory) neural network architecture to recognize text. Users can interact with Tesseract directly through the command line or integrate it into their own software applications using its fully featured API. The engine relies on external data files (traineddata) to support different languages and scripts, which can be swapped or updated depending on the specific recognition task.

Who it’s for

Tesseract is designed for developers who need to build text-recognition capabilities into their own software, as well as power users comfortable with command-line tools. Because it is highly portable, it can be compiled for various platforms, including Android and iPhone, making it suitable for both desktop and mobile applications.

Highlights

  • Supports over 100 languages and 35 scripts.
  • Offers a flexible API for custom software integration.
  • Provides multiple model options, including 'fast' and 'best' versions, to balance speed and accuracy.
  • Highly portable, with support for Linux, Windows, Android, and iOS.
  • Open-source software licensed under the Apache 2.0 license.

Related

  • Project
    schappim/macOCR

    A command-line OCR tool for macOS that extracts text, QR codes, and barcodes from the screen, images, or PDFs using Apple's Vision framework.

  • Project
    wolny/pytorch-3dunet

    pytorch-3dunet provides PyTorch implementations of 3D U-Net and its variants (Residual, Squeeze-and-Excitation) for semantic segmentation and regression tasks on volumetric data, with command-line training and prediction tools, multiple loss functions and evaluation metrics.

  • Project
    twistedfall/opencv-rust

    Rust bindings for the OpenCV computer vision library, enabling the use of OpenCV's C++ API within Rust applications.

  • Project
    b7leung/MLE-Flashcards

    A 250+ flashcard deck covering ML topics (CS, classical ML, deep learning, vision, NLP, RL, generative models), intended for people with a solid ML foundation. Updated May 2025 with newer topics like LLMs and NeRFs.