OCR libraries for C#?

Question

C++ / C#

Tried IronOCR and Tesseract OCR (wrapper for .NET). In Tesseract, I tried different page fragmentation modes - the result is bad - garbage or part of the text and garbage. The input is a binarized ( B / W ) image of the document. There is not only text in the photo - there is also rubbish in the form of a photograph and a person’s signature, but I will try to eliminate it. Another difficulty is that the image contains several languages ​​at the same time - Russian and English or English and Farsi (Arabic). So far I'm trying to deal with Russian and English - at least. Are there any other libraries / alternatives - to the above?