PARAGON: Intelligent Digital Collation and Difference Detection

FAIN: HK-50032-12

University of South Carolina Research Foundation (Columbia, SC 29208-0001)
David Lee Miller (Project Director: January 2012 to May 2017)

The further development of advanced collation and difference detection software that allows for the robust comparison of scanned images of texts, even if those scans include errors like warping of the page.

PARAGON builds on a successful prototype funded by an NEH startup grant in 2009. This prototype demonstrated that combining SIFT (scale-invariant feature transform) and TPS (Thin-Plate Spline) algorithms makes it possible to automate the detailed comparison of scanned images captured under varying circumstances, whether scanned, camera-taken from different heights or angles, rotated, differently lighted, or even slightly warped. PARAGON will develop this prototype into a robust, open-source software capable of analyzing and collating large sets of digital images.

Media Coverage

Burns in the heat of the South: rare books and modern technology (Review)
Author(s): Craig Lamont
Publication: Electric Scotland
Date: 9/14/2016
Abstract: The author is a Robert Burns scholar who describes visiting the University of South Carolina to consult materials held in Special Collections here. He talks about experimenting with Paragon in the work of collating Burns materials.
URL: http://www.electricscotland.com/familytree/frank/burns_lives243.htm

Associated Products