tesseractgt: Generate Ground Truth Data for Tesseract

In an early stage (v0.0.4). Check out the blog post with a fully developed example and the package documentation.

Step-by-step summary

(Go to the blog post for more detailed explanations of each step.)

0. Install the Tesseract engine, tesseract, and tesseractgt.

1. Take screenshots to serve as fine-tuning images, and save them in a folder.

2. Use tesseractgt::create_gt_txt() to create and pre-fill files with the text in the images.

3. Call the GUI from "Addins > Correct ground truth files" in RStudio or tesseractgt::correct_gt_txt(). Use it to correct the text files created in step 2.

4. Fine-tune using tesstrain.

Installation

install.packages("remotes") # if `remotes` is not installed
remotes::install_github("arcruz0/tesseractgt")