Verifying the integrity of hardcopy document using OCR

Abstract

Hardcopy document forgery is still a challenge and occurs frequently nowadays. Many countries have reported a lot of cases, including South Africa where government issues documents are forged. Protecting these documents from being tampered with is necessary at all times. Various methods have been presented to deal with the challenge of document forgery such as, e.g. Optical Character Recognition (OCR). In this paper, we improve OCR with the aim to achieve a high accuracy to eliminate the misrepresentation of characters read from an image file. To implement the solution we use an OCR tool, Tesseract. The experimental setup is explained and the results which yielded an accuracy of 100% are discussed in detail. While this is on-going work, the experimental results demonstrate the feasibility of using OCR as part of the solution.