OCR translation still isn’t perfect but it has improved dramatically over the past few years. Leading the way is the Tesseract translation engine currently open sourced in C++. While this is an incredible library, it’s however, restricted to software.
Full article