An Automated Technique to Recognize and Extract Images from Scanned Archaeological Documents

Download PDF
Open Webpage

Cindy Roullet, David Fredrick, John Gauch, Rhodora Vennarucci

Published: 01 Jan 2019, Last Modified: 07 Nov 2025GREC@ICDAR 2019EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Pompei: pitture e mosaici" is a valuable set of volumes containing over 20,000 historical annotated images of the archaeological site of Pompeii, Italy. Our project consists of extracting, archiving, analyzing and classifying all the image data from a digitized version of these books. In this paper, we describe a method that automatically locates and separates graphical elements such as maps, drawings, paintings and photographic images from text. We also introduce our ongoing work on the interpretation of the retrieved data.