Russian text recognition from scanned images, how to implement?

Question

linux

There is a task to recognize Russian text from scanned images / documents (*.jpg or *.pdf) and then enter this data into the database.
The question is, who had the experience of such development and by what means is it better to implement it?
PS if there are libraries similar to PHP, then this is the preferred solution for me as a beginner)
Thanks in advance for your advice and answers!

Answer 1

tesseract is from google, cuneiform is there
.

Answer 2

This is definitely not done in php, because the resource-intensive php operation is too slow for this.