DEEP LEARNING TEXT RECOGNITION AND EXTRACTION ON FORMATTED BILLS
CENTRAL ASIAN JOURNAL OF MATHEMATICAL THEORY AND COMPUTER SCIENCES, 4 (2):
1-18 (February 2023)Abstract
Image processing, including the processing of invoices and handwritten bills, has become indispensable in all areas of business today. The main drawback is having to manually retype the material from invoices. As a result, extracting text from photos in their natural settings remains a formidable challenge. This study provides an in-depth look at text recognition for the purposes of extracting data from commonly used bill and invoice formats, such as handwritten notes and printed documents, for the purpose of automating database updates. The method developed in this study provides a quick and reliable means of extracting text from the bill. In this study, we will combine deep learning with Optical Character Recognition to analyse text in images and scanned documents and convert it into machine-readable text. After that, we use OpenCV to pull the text from the image, and then run it through our algorithm to ensure the highest level of accuracy possible before automatically saving the new text to our database.