Document Image Extraction System Design

Abstract

Abstract The design of the document image extraction system aims to provide an overview of the process in a system that converts the image of a document into text so that it is easier to use (save, manage or search for information) or representing in interesting visualization. The system design based on specific cases such as decision letters, certificates, and assignments. The document identified data that might be needed. After that, other processes are carried out based on each document. The design results show the flow of the system starts from data input, scanning, pre-processing to the document image, character classification, normalization process, and extraction process. This is done because the uniqueness of each document requires a unique process so that it cannot be treated in general. The design of the document image extraction system focuses more on the process of character recognition. The uniqueness of a document has an impact on the extraction process. Then the system design needs to be added to selected the type of document to be extracted.