Document segmentation using deep learning

Abstract

In recent years, with the rise of computational power and the growing popularity of artificial intelligence, the fields in data analysis have increased significantly. Big leaps forwards have been made with regards to object detection such as locating a receipt in an image and in computer vision. However, data analysis for unstructured data such as parsing text from a receipt remains a challenge and little research has been made. Receipts are especially challenging since there is no standard format for text placement or keywords such as total which differ from on evendor to another. This dissertation will discuss and explore an image-based and a text-based implementation in order to extract key details such as Shop name, Date of the receipt, VAT number and total from a receipt. The image-based model yielded no results in its current state and acts as a proof of concept and with enough time and data, it could be a viable solution in the future. On the other hand, the text-based model has managed to yield promising results. The tests conducted include a comparison of this model with two other existing products on the market and the results are considered a success. With this model, the groundwork has been made and with further training and optimization, this solution has the potential to rival the market’s solutions.