The goal of this project is to be able to convert
into these two pages :
The output is a PDF with final images with OCR behind the image.
Now, the process can split pages in two, rotate each page, remove black border and save them to individual images.
Write all documentation at the same time that coding is hard. Some algorithms are too simply and I know they need to be improved.
You can have a look at the file run_all.py and simply run the command execute(path_to_image)
No doubt about it. It still the beginning. Open a issue in github, send the image and tell what the problem. I will do my best to improve the recognition.
Open the file algorithm.asciidoc with the browser but remember that full algorithm is not detailed until they are stable.