I'd like to explore the possibilities of applying artificial intelligence to ocr reading. Basic ocr invoices processing let me convert 30% of them only. The main purpose is defining invoices areas by training an ai, then process those areas with ocr. So I am looking into ai to define and recognize a document topology first, then apply ocr locally. From a brief search, it is classified as zonal or template ocr. Any chance of a premade open source library?