tobiasmcvey / learn-ocr

examples using tesseract and opencv in Norwegian (Norsk bokmål) and English

Notes from learning OCR

Examples using Tesseract and OpenCV to convert pictures to text.

Installing tesseract adds support for English. To support additional languages, install it separately or install all available languages.

introduction.py is an introduction script to show how to parse text from an image with interactive python.

oppskrifter.py is an example of converting pictures with recipes in Norwegian to text files and searchable PDF files.

vildanden.py is an example of converting a picture of the starting act of Vildanden by Henrik Ibsen, into a text file and a searchable PDF. See the image from Vildanden on Project Gutenberg provided under Public Domain in the USA.

Setup

On linux

sudo apt-get update
sudo apt-get install tesseract-ocr
sudo apt-get install libtesseract-dev

and install dependencies in requirements file with pip install -r requirements.txt

converted .jpg files to .png for easier conversion using ffmpeg using ffmpeg -i input.jpg output.png