Self-verifiable Paper Document and Automatic Content Authentication
Yibin Tian, Gang Fang, Ming Wei · Konica-Minolta Technology Report · 2014
This report describes a solution for self-verifiable paper documents. The content of the original document is extracted and encoded in high-capacity color barcodes, which are printed out on the same paper as the original image to form an authentic document. For content authentication, the document is scanned to obtain the barcodes and a 2nd image. The original image is reconstructed from barcode data, and compared to the 2nd image using multiple features. It uti lizes regular office computers, multi-functional printers and papers, and supports multiple languages without using OCR. Thus it is inexpensive, robust, and more versatile compared to existing solutions. Evaluation on 353 documents in multiple Western languages scanned at 600 dpi achieved greater than 99 % alteration detection rate and less than 0.9 % false positive rate at the word/symbol level, in 25s per page.