Theres a series of Optical Character Recognition repos that should help you with task #1. They are all based around Google's Tesseract. If I remember correctly this is one of the top=> https://github.com/danielquinn/paperless I've used project naptha in the past... and little known fact that google docs can do the OCR automatically too.

regarding the translation... never had to do it. sorry!