Extracting information from printed grocery receipts is a promising area of research for many years but yet researched in very limited situations. Receipt image optimization, Receipt image to text transformation and information extraction are used as the front end approach for the development of Receipt processing. Sequential based information extraction is cutting edge approach in semi structure domain. In this paper, we provide an approach to extract information from printed grocery receipts. We demonstrated F1 score of nearly 1 for single supermarket and we performed error analysis on template independent information extraction from grocery receipts.
Team Members: Chuhaashanan Nagenthiean
