This repo contains an algorithm that is able to extract the content of receipts. It uses the OCR algorithm of Google Vision API to parse text from scanned images of the recepits.
The algorithm is rule-based and is constructed from Swedish supermarket receipts. Therefore, it may not work very well for your receipts, if they are very different from these.