This repository contains a dataset of labeled and un-labeled images which is an extension of DSBI and AngelinaDataset.
The dataset will be used for creation of a Braille characters detector based on the AngelinaReader. The current version of the detector is here: braille-systems/AngelinaReader
Git with Git LFS
To find out whether you have Git with LFS installed, run
git lfs --version
To checkout the files:
git clone --recursive https://github.com/braille-systems/brl_ocr.git
Labels for data are made with labelme. To view, add or edit them, install and run this tool (preferrably in a fresh pip environment):
python -m pip install labelme
labelme --config .labelmerc
Open the desired picture, e. g. brl_ocr/data/labeled/books/golubina/IMG_2504.JPG:
